notebooks/experiments/02-compare_annotations_Herbold.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "468fcf1e-7ed4-48f7-a303-89ddcb61e6ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import re\n",
    "import stat\n",
    "from collections import Counter, defaultdict\n",
    "from pathlib import Path\n",
    "from textwrap import dedent\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "from bson import json_util\n",
    "from bson.objectid import ObjectId\n",
    "from mongoengine import connect, disconnect, DoesNotExist\n",
    "from pycoshark.mongomodels import Commit, FileAction, File, Project, VCSSystem, Hunk, Issue, IssueSystem, Refactoring\n",
    "from pycoshark.utils import create_mongodb_uri_string, java_filename_filter"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60a74919-7493-47c7-86fc-4d4b3b39552a",
   "metadata": {},
   "source": [
    "# Comparison with line annotations in the Herbold et al. dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb30b547-45ed-438f-9111-91567bcb85fa",
   "metadata": {},
   "source": [
    "- S. Herbold, A. Trautsch, B. Ledel, A. Aghamohammadi, T. A. Ghaleb, K. K. Chahal, T. Bossenmaier, B. Nagaria, P. Makedonski, M. N. Ahmadabadi, K. Szabados, H. Spieker, M. Madeja, N. Hoy, V. Lenarduzzi, S. Wang, G. Rodríguez-Pérez, R. Colomo-Palacios, R. Verdecchia, P. Singh, Y. Qin, D. Chakroborti, W. Davis, V. Walunj, H. Wu, D. Marcilio, O. Alam, A. Aldaeej, I. Amit, B. Turhan, S. Eismann, A.-K. Wickert, I. Malavolta, M. Sulír, F. Fard, A. Z. Henley, S. Kourtzanidis, E. Tuzun, C. Treude, S. M. Shamasbi, I. Pashchenko, M. Wyrich,\n",
    "J. Davis, A. Serebrenik, E. Albrecht, E. U. Aktas, D. Strüber, and J. Erbel, _“A fine-grained data set and analysis of tangling in bug fixing\n",
    "commits,”_ Empirical Software Engineering, vol. 27, no. 6, p. 125, Nov. 2022<br>https://doi.org/10.1007/s10664-021-10083-5\n",
    "- S. Herbold, A. Trautsch, B. Ledel, _“Large-Scale Manual Validation of Bugfixing Changes.”_ In Proceedings of the 17th International Conference on Mining Software Repositories (MSR '20). 2020. Association for Computing Machinery, New York, NY, USA, 611–614.<br> https://doi.org/10.1145/3379597.3387504"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "80849ac7-1da9-4855-904b-aea3147ec540",
   "metadata": {},
   "source": [
    "Replication package: \n",
    "- https://github.com/sherbold/replication-kit-2020-line-validation (last commit from Jul 30, 2021, 13 stars, 5 forks)\n",
    "- https://github.com/smartshark/usage-examples (last commit from Aug 3, 2021, 6 starts 4 forks)\n",
    "\n",
    "Release of Replication Kit (`replication-kit-2020-line-validation`) on May 6, 2022 - tagging<br>\n",
    "https://github.com/sherbold/replication-kit-2020-line-validation/releases"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e8f6adb-b115-43e4-9965-77519be3b23d",
   "metadata": {},
   "source": [
    "Python data collection library for SmartSHARK: https://github.com/smartshark/pycoSHARK\n",
    "\n",
    "Used in `Replication-Notebook.ipynb`:\n",
    "\n",
    "```python\n",
    "from pycoshark.mongomodels import Commit, FileAction, File, Project, VCSSystem, Hunk, Issue, IssueSystem, Refactoring\n",
    "from pycoshark.utils import create_mongodb_uri_string, java_filename_filter\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "467977b3-2be3-4f85-8eaf-28c021bc66df",
   "metadata": {},
   "source": [
    "> **Contents of the repository \\[with replication package]**\n",
    "> - The `Replication-Notebook.ipynb` with all code required to reproduce our results from the raw data.\n",
    "> - The data folder with the `hunk_labels.json` and the `leaderboard.json` file.\n",
    ">     - The `hunk_labels.json` contains the relevant raw data for this study, i.e., the manual labels for the hunks.\n",
    ">     - The `leaderboard.json` contains daily snapshots of the progress per user and per project.\n",
    "> - The `figures/` folder with all result figures that are generated by the Replication-Notebook."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0d72e2c-0c43-4d0f-9186-ebd2856f1a85",
   "metadata": {},
   "source": [
    "## Reading 'hunk_labels.json' for hunk_id's"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ff5508f-5546-4747-a662-96f48a0e900b",
   "metadata": {},
   "source": [
    "We will be using `'hunk_labels.json'` file to limit to those anotations that were considered in Herbold et al. 2022 paper.\n",
    "\n",
    "The MongoDB database that we will be using to augment the serialized data with information necessary to be able to compare with PatchScope (`diff-annotate`) annotations is `'smartshark_2_2'`, while `Replication-Notebook.ipyb` uses `'smartshark_1_2'`\n",
    "\n",
    "At the time of writing this, the most recent version of SmartSHARK database is \"SmartSHARK MongoDB Release 2.2\"<br>\n",
    "See https://smartshark.github.io/dbreleases/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a381cfc4-1c5d-488e-aa58-d40b09a46396",
   "metadata": {},
   "outputs": [],
   "source": [
    "# configuration local to the computer the notebook is being run on (!!!)\n",
    "smartshark_dir = '/mnt/data/msr/smartshark_repositories/replication-kit-2020-line-validation'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "d1698b29-bd9c-4b18-9b58-10cb5bcf5ea0",
   "metadata": {},
   "outputs": [],
   "source": [
    "hunk_data_file = 'data/hunk_labels.json'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "fd510a5e-661e-4e89-9891-dd072ce6e26a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# check that the path is correct\n",
    "hunk_data_path = Path(smartshark_dir).joinpath(hunk_data_file)\n",
    "assert hunk_data_path.is_file()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "356916df-0e48-4418-bfd7-fd37ae5a7032",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "list"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read the data\n",
    "with open(hunk_data_path, mode='r') as json_fp:\n",
    "    smartshark_data = json.load(json_fp)\n",
    "\n",
    "type(smartshark_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b5da8c3-80fd-46d2-97b6-30ebf7a142e8",
   "metadata": {},
   "source": [
    "You can find detailed examination of the `'hunk_labels.json'` file schema in [_\"Examining the format of SmartSHARK's 'hunk_labels.json'\"_](./01-compare_annotations.ipynb#Examining-the-format-of-SmartSHARK's-'hunk_labels.json') seection of the [`01-compare_annotations.ipynb`](./01-compare_annotations.ipynb) Jupyter Notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a5c51c6a-98b8-4c82-bb3d-f4407a07ba32",
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_smartshark_hunk_ids(hunk_labels_data: list[dict]) -> list[str]:\n",
    "    return [\n",
    "        entry['hunk_id']['$oid'] for entry in hunk_labels_data\n",
    "    ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a7611134-c824-4fbd-928a-3391c61a31a6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['5a82f5aa912063217b88940a',\n",
       " '5a82f5aa912063217a8893ef',\n",
       " '5a82f5aa912063217b88940c',\n",
       " '5a82f5aa912063217a8893f1',\n",
       " '5a82f5aa912063217b88940e']"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_hunk_ids = extract_smartshark_hunk_ids(smartshark_data)\n",
    "smartshark_hunk_ids[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "0f60f661-e14c-4830-9233-fc18225bbdcc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "len(smartshark_data)=31065\n",
      "len(smartshark_hunk_ids)=31065\n"
     ]
    }
   ],
   "source": [
    "print(f\"{len(smartshark_data)=}\")\n",
    "print(f\"{len(smartshark_hunk_ids)=}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "8ab4b012-f263-4c06-a5ed-ba61a00cb444",
   "metadata": {},
   "outputs": [],
   "source": [
    "smartshark_hunk_ids_set = set(smartshark_hunk_ids)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ce721b8f-f3e9-403b-9f39-40db3ebe3d13",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43.4 ns ± 0.236 ns per loop (mean ± std. dev. of 7 runs, 10,000,000 loops each)\n",
      "22.3 ns ± 0.047 ns per loop (mean ± std. dev. of 7 runs, 10,000,000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%timeit '5a82f5aa912063217b88940c' in smartshark_hunk_ids\n",
    "%timeit '5a82f5aa912063217b88940c' in smartshark_hunk_ids_set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13dcb50a-7812-4684-a73c-69d7af1594c1",
   "metadata": {},
   "source": [
    "## Load data from SmartSHARK MongoDB database"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0bea2030-b054-4464-9498-fdf78f2d5d0f",
   "metadata": {},
   "source": [
    "The code is based on [`Replication-Notebook.ipynb`](https://github.com/sherbold/replication-kit-2020-line-validation/blob/main/Replication-Notebook.ipynb) in [sherbold/replication-kit-2020-line-validation](https://github.com/sherbold/replication-kit-2020-line-validation) repository.\n",
    "\n",
    "That code uses [mongoengine](https://pypi.org/project/mongoengine/) (an ORM-like layer on top of PyMongo, a Python Object-Document Mapper for working with MongoDB) and [pycoSHARK](https://pypi.org/project/pycoSHARK/) (basic MongoDB models for smartSHARK) packages.\n",
    "\n",
    "We could have used only [PyMongo](https://pypi.org/project/pymongo/), but those packages mentioned above do offer nice conveniences.\n",
    "\n",
    "```python\n",
    "from pymongo import MongoClient\n",
    "\n",
    "client = MongoClient(host, port)\n",
    "\n",
    "smartshark_2_2_db = client['smartshark_2_2']\n",
    "commits_table = smartshark_2_2_db['commit']\n",
    "file_table = smartshark_2_2_db['file']\n",
    "file_action_table = smartshark_2_2_db['file_action']\n",
    "hunk_table = smartshark_2_2_db['hunk']\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "eca8e83e-608f-41d3-9f0e-271881e03929",
   "metadata": {},
   "outputs": [],
   "source": [
    "# configuration local to the computer the notebook is being run on (!!!)\n",
    "credentials = {\n",
    "    'db_user': '',\n",
    "    'db_password': '',\n",
    "    'db_hostname': 'localhost',\n",
    "    'db_port': 37017,  # standard port is 27017\n",
    "    'db_authentication_database': '',\n",
    "    'db_ssl_enabled': False\n",
    "}\n",
    "\n",
    "database_name = 'smartshark_2_2'  # replication package for Herbold uses 'smartshark_1_2'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "c63260a6-775f-4e90-8516-ded88b4821a7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'mongodb://localhost:37017'"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "uri = create_mongodb_uri_string(**credentials)\n",
    "uri"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "e2e97633-c21f-4f29-8a96-f4a32ef8317a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MongoClient(host=['localhost:37017'], document_class=dict, tz_aware=False, connect=True, read_preference=Primary())"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "client = connect(database_name, host=uri, alias='default')\n",
    "client"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e535cf33-76cb-472a-8003-434e80ec6b61",
   "metadata": {},
   "source": [
    "Check that everything works correctly, by retrievieng example \"hunk\", with `$oid` of `'5a82f5ab912063217d88943b'`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "07f0ae57-76cb-423a-9922-ab11fc6d185d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SON([('_id', ObjectId('5a82f5ab912063217d88943b')), ('file_action_id', ObjectId('58c898ebc03bb4351fbcad44')), ('new_start', 110), ('new_lines', 4), ('old_start', 109), ('old_lines', 2), ('content', '-        BufferedReader r = new BufferedReader(new InputStreamReader(URLHandlerRegistry.getDefault()\\n-                .openStream(url)));\\n+        URLHandler urlHandler = URLHandlerRegistry.getDefault();\\n+        String charset = urlHandler.getURLInfo(url).getBodyCharset();\\n+        InputStream contentStream = urlHandler.openStream(url);\\n+        BufferedReader r = new BufferedReader(new InputStreamReader(contentStream, charset));\\n'), ('lines_manual', {'sherbold': {'bugfix': [0, 1, 2, 3, 4, 5]}, 'riruk': {'refactoring': [0, 1, 4, 5], 'bugfix': [2, 3]}, 'aaghamohammadi': {'bugfix': [0, 1, 2, 3, 4, 5]}, 'omalam': {'refactoring': [0, 1, 4, 5], 'bugfix': [2, 3]}}), ('lines_verified', {'None': [0, 1, 4, 5], 'bugfix': [2, 3]}), ('uncorrected_lines_manual', {})])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Hunk.objects(pk='5a82f5ab912063217d88943b').first().to_mongo()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "fe53933f-1123-4354-ac6b-f80cfa738792",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ObjectId('5a82f5ab912063217d88943b')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h_id = Hunk.objects(pk='5a82f5ab912063217d88943b').first().id\n",
    "h_id"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "d4737998-7c43-454d-8d95-90cac43e7fd9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'5a82f5ab912063217d88943b'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "str(h_id)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "eafda194-861c-4929-a356-8585c61d62f9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'sherbold': {'bugfix': [0, 1, 2, 3, 4, 5]},\n",
       " 'riruk': {'refactoring': [0, 1, 4, 5], 'bugfix': [2, 3]},\n",
       " 'aaghamohammadi': {'bugfix': [0, 1, 2, 3, 4, 5]},\n",
       " 'omalam': {'refactoring': [0, 1, 4, 5], 'bugfix': [2, 3]}}"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Hunk.objects(pk='5a82f5ab912063217d88943b').first().lines_manual"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "4ef56509-d269-41cc-b313-5743ba36680c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'None': [0, 1, 4, 5], 'bugfix': [2, 3]}"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Hunk.objects(pk='5a82f5ab912063217d88943b').first().lines_verified"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "92f8bbc3-5e26-4f0d-948b-0265088cd88d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['-        BufferedReader r = new BufferedReader(new InputStreamReader(URLHandlerRegistry.getDefault()',\n",
       " '-                .openStream(url)));',\n",
       " '+        URLHandler urlHandler = URLHandlerRegistry.getDefault();',\n",
       " '+        String charset = urlHandler.getURLInfo(url).getBodyCharset();',\n",
       " '+        InputStream contentStream = urlHandler.openStream(url);',\n",
       " '+        BufferedReader r = new BufferedReader(new InputStreamReader(contentStream, charset));']"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Hunk.objects(pk='5a82f5ab912063217d88943b').first().content.split('\\n')[:-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f4782b5-99a7-4671-aea2-89e09ed06c91",
   "metadata": {},
   "source": [
    "-----\n",
    "\n",
    "Exploration for multi-group hunk\n",
    "\n",
    "We need to find line number in file, in pre-image or post-image, accordingly.  Multi-group hunk represents the most difficult case."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "1a92fb5b-db87-4279-8985-5cf446745d85",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['-   public static TokenNameFinderModel train(String languageCode, String type, ObjectStream<NameSample> samples, ',\n",
       " '-       byte[] generatorDescriptor, final Map<String, Object> resources, ',\n",
       " '+   public static TokenNameFinderModel train(String languageCode, String type, ObjectStream<NameSample> samples,',\n",
       " '+       byte[] generatorDescriptor, final Map<String, Object> resources,',\n",
       " '        int iterations, int cutoff) throws IOException {',\n",
       " '-     ',\n",
       " '+',\n",
       " '      // TODO: Pass in resource manager ...',\n",
       " '-     ',\n",
       " '+',\n",
       " '      AdaptiveFeatureGenerator featureGenerator = createFeatureGenerator(generatorDescriptor, resources);',\n",
       " '-     ',\n",
       " '+']"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hunk = Hunk.objects(pk='5b27c661eee553716977ddcb').first()\n",
    "\n",
    "hunk_lines_data = {}\n",
    "hunk_lines = hunk.content.split('\\n')[:-1]  # remove trailing empty element, because '\\n' is terminator\n",
    "hunk_count_m = 0\n",
    "hunk_count_p = 0\n",
    "hunk_offsets_m = {}\n",
    "hunk_offsets_p = {}\n",
    "for i, line in enumerate(hunk_lines):\n",
    "    line_type = line[:1]\n",
    "    if line_type == '-':\n",
    "        hunk_offsets_m[i] = hunk_count_m\n",
    "        hunk_count_m += 1\n",
    "    elif line_type == '+':\n",
    "        hunk_offsets_p[i] = hunk_count_p\n",
    "        hunk_count_p += 1\n",
    "    elif line_type == ' ':\n",
    "        hunk_count_m += 1\n",
    "        hunk_count_p += 1\n",
    "    else:\n",
    "        pass\n",
    "\n",
    "hunk_lines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "21570244-22e1-47e8-b095-d6a9c2883105",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0: 0, 1: 1, 5: 3, 8: 5, 11: 7}"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hunk_offsets_m"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "189e33b3-d7e2-421c-aef3-e81a62f69adb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['-   public static TokenNameFinderModel train(String languageCode, String type, ObjectStream<NameSample> samples, ',\n",
       " '-       byte[] generatorDescriptor, final Map<String, Object> resources, ',\n",
       " '-     ',\n",
       " '-     ',\n",
       " '-     ']"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[hunk_lines[off] for off in hunk_offsets_m.keys()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "811e68e0-b3e7-4da7-94ef-abca0ea493f4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 0 |  0:-   public static TokenNameFinderModel train(Strin |   :                                                  \n",
      " 1 |  1:-       byte[] generatorDescriptor, final Map<Stri |   :                                                  \n",
      " 2 |   :                                                   |  0:+   public static TokenNameFinderModel train(Strin\n",
      " 3 |   :                                                   |  1:+       byte[] generatorDescriptor, final Map<Stri\n",
      " 4 |  _:        int iterations, int cutoff) throws IOExcep |  _:        int iterations, int cutoff) throws IOExcep\n",
      " 5 |  3:-                                                  |   :                                                  \n",
      " 6 |   :                                                   |  3:+                                                 \n",
      " 7 |  _:      // TODO: Pass in resource manager ...        |  _:      // TODO: Pass in resource manager ...       \n",
      " 8 |  5:-                                                  |   :                                                  \n",
      " 9 |   :                                                   |  5:+                                                 \n",
      "10 |  _:      AdaptiveFeatureGenerator featureGenerator =  |  _:      AdaptiveFeatureGenerator featureGenerator = \n",
      "11 |  7:-                                                  |   :                                                  \n",
      "12 |   :                                                   |  7:+                                                 \n"
     ]
    }
   ],
   "source": [
    "w = 50\n",
    "for i, line in enumerate(hunk_lines):\n",
    "    line_type = line[:1]\n",
    "    l = \"\" if line_type == '+' else line\n",
    "    r = \"\" if line_type == '-' else line\n",
    "    \n",
    "    ln = \"\"\n",
    "    rn = \"\"\n",
    "    if line_type == '-':\n",
    "        if i in hunk_offsets_m:\n",
    "            ln = \"{n:2d}\".format(n=hunk_offsets_m[i])\n",
    "        else:\n",
    "            ln = \"??\"\n",
    "    if line_type == '+':\n",
    "        if i in hunk_offsets_p:\n",
    "            rn = \"{n:2d}\".format(n=hunk_offsets_p[i])\n",
    "        else:\n",
    "            rn = \"??\"\n",
    "    if line_type == ' ':\n",
    "        ln = \"_\"\n",
    "        rn = \"_\"\n",
    "    \n",
    "    print(f\"{i:2d} | {ln:>2s}:{l:<{w}.{w}s} | {rn:>2s}:{r:<{w}.{w}s}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "4ff22d7e-5546-4e37-b20c-5e59a27bdbed",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 0 | 450:-   public static TokenNameFinderModel train(Strin |    :                                                  \n",
      " 1 | 451:-       byte[] generatorDescriptor, final Map<Stri |    :                                                  \n",
      " 2 |    :                                                   | 450:+   public static TokenNameFinderModel train(Strin\n",
      " 3 |    :                                                   | 451:+       byte[] generatorDescriptor, final Map<Stri\n",
      " 4 | ___:        int iterations, int cutoff) throws IOExcep | ___:        int iterations, int cutoff) throws IOExcep\n",
      " 5 | 453:-     ⋅                                            |    :                                                  \n",
      " 6 |    :                                                   | 453:+⋅                                                \n",
      " 7 | ___:      // TODO: Pass in resource manager ...⋅       | ___:      // TODO: Pass in resource manager ...⋅      \n",
      " 8 | 455:-     ⋅                                            |    :                                                  \n",
      " 9 |    :                                                   | 455:+⋅                                                \n",
      "10 | ___:      AdaptiveFeatureGenerator featureGenerator =  | ___:      AdaptiveFeatureGenerator featureGenerator = \n",
      "11 | 457:-     ⋅                                            |    :                                                  \n",
      "12 |    :                                                   | 457:+⋅                                                \n"
     ]
    }
   ],
   "source": [
    "w = 50\n",
    "for i, line in enumerate(hunk_lines):\n",
    "    line_type = line[:1]\n",
    "    l = \"\" if line_type == '+' else line+'⋅'\n",
    "    r = \"\" if line_type == '-' else line+'⋅'\n",
    "    \n",
    "    ln = \"\"\n",
    "    rn = \"\"\n",
    "    if line_type == '-':\n",
    "        if i in hunk_offsets_m:\n",
    "            ln = \"{n:3d}\".format(n=hunk_offsets_m[i] + hunk.old_start)\n",
    "        else:\n",
    "            ln = \"??\"\n",
    "    if line_type == '+':\n",
    "        if i in hunk_offsets_p:\n",
    "            rn = \"{n:3d}\".format(n=hunk_offsets_p[i] + hunk.new_start)\n",
    "        else:\n",
    "            rn = \"??\"\n",
    "    if line_type == ' ':\n",
    "        ln = \"___\"\n",
    "        rn = \"___\"\n",
    "    \n",
    "    print(f\"{i:2d} | {ln:>3s}:{l:<{w}.{w}s} | {rn:>3s}:{r:<{w}.{w}s}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "83b3b82c-c15e-4070-85e5-0ee8ed57fab6",
   "metadata": {},
   "source": [
    "The above is not _quite_ the side by side diff view, but close enough that one can check if the code for computing line numbers (line number in pre-image for '-' lines, line number in post-image for '+' lines) produces correct results.\n",
    "\n",
    "Compare with side-by-side diff in GitHub:<br>\n",
    "https://github.com/apache/opennlp/commit/4496b1e746451527bfb05e7ea48358c4ae509e1b"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "81afa1bb-1513-4b61-b2cd-d69a42cf2b19",
   "metadata": {},
   "source": [
    "End of exploration\n",
    "\n",
    "-----"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a889d390-d3c2-4783-8e27-dfbb301a1beb",
   "metadata": {},
   "source": [
    "Code based on the code in  `Replication-Notebook.ipynb` in sherbold/replication-kit-2020-line-validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "7f1e8f66-2c74-48e6-a5b5-bfc8b790e3dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# switch to False if you want to avoid queruing the MongoDB database\n",
    "use_mongodb = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "3abbdaa0-e29e-4d6f-b22e-14d2b9dc59ba",
   "metadata": {},
   "outputs": [],
   "source": [
    "# set to true for debugging issues\n",
    "is_test = False"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6e18295-246c-40c5-9f47-d79c34dcdb07",
   "metadata": {},
   "source": [
    "Running with `use_mongodb` set to `True` takes around 8min 15s\n",
    "- running without `is_test` set to `False` takes 4.5 minutes to only load list of hunks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "ef6ea573-78f6-48fa-9a81-04288baad558",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "len(smartshark_hunk_ids)=31065 (present in 'hunk_labels.json')\n",
      "len(hunks_extra)=5813357 (outside 'smartshark_1_2' used in Herbold et al., and present in 'hunk_labels.json')\n",
      "len(hunks_miss)=0 (with 3 or less annotators/users)\n",
      "len(hunks_multi_issue)=0 (commits that link to multiple bugs)\n",
      "len(smartshark_records)=296348 lines (results of parsing)\n",
      "\n",
      "CPU times: user 5min 55s, sys: 17.8 s, total: 6min 12s\n",
      "Wall time: 7min 14s\n",
      "Parser   : 120 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "# NOTE that using %%time inhibits debugging errors\n",
    "\n",
    "if use_mongodb:\n",
    "    smartshark_records = []\n",
    "    hunks_extra = []\n",
    "    hunks_miss = []\n",
    "    hunks_multi_issue = []\n",
    "\n",
    "    if is_test:\n",
    "        # for faster testing, run process for single hunk\n",
    "        hunks = [Hunk.objects(pk='5b27c661eee553716977ddcb').first()]\n",
    "    else:\n",
    "        # cache hunks locally to avoid timeouts\n",
    "        hunks = [h for h in Hunk.objects(lines_manual__exists=True)]\n",
    "\n",
    "    for hunk in hunks:\n",
    "        hunk_id = str(hunk.id)\n",
    "        \n",
    "        if hunk_id not in smartshark_hunk_ids_set:\n",
    "            hunks_extra.append(hunk_id)\n",
    "            continue\n",
    "        if len(hunk.lines_manual) <= 3:\n",
    "            hunks_miss.append(hunk_id)\n",
    "            continue\n",
    "\n",
    "        fa = FileAction.objects(id=hunk.file_action_id).get()\n",
    "        file = File.objects(id=fa.file_id).get()\n",
    "        commit = Commit.objects(id=fa.commit_id).only('revision_hash', 'fixed_issue_ids', 'vcs_system_id').get()\n",
    "        vcs = VCSSystem.objects(id=commit.vcs_system_id).get()\n",
    "        project = Project.objects(id=vcs.project_id).get()\n",
    "        \n",
    "        external_id = None\n",
    "        num_fixed_bugs = 0\n",
    "        for issue in Issue.objects(id__in=commit.fixed_issue_ids):\n",
    "            if issue.issue_type_verified is not None and issue.issue_type_verified.lower() == 'bug':\n",
    "                num_fixed_bugs += 1\n",
    "                external_id = issue.external_id\n",
    "\n",
    "        if num_fixed_bugs != 1:\n",
    "            hunks_multi_issue.append({'hunk_id': hunk_id, 'num_fixed_bugs': num_fixed_bugs})\n",
    "            continue\n",
    "\n",
    "        hunk_info = {\n",
    "            'project': project.name,\n",
    "            'repository_url': vcs.url,\n",
    "            'issue_id': issue.external_id,\n",
    "            'revision_hash':commit.revision_hash,\n",
    "            'file': file.path,\n",
    "            'hunk_id': hunk_id,\n",
    "            'old_start': hunk.old_start,\n",
    "            'old_lines': hunk.old_lines,\n",
    "            'new_start': hunk.new_start,\n",
    "            'new_lines': hunk.new_lines,\n",
    "        }\n",
    "\n",
    "        # in chunk of changes/change group, removed lines/old lines ('-') always precede added lines/new lines ('+')\n",
    "        # but \"hunk\" in SmartSHARK data does not need to consist of single change group,\n",
    "        # see hunk with id '5b27c661eee553716977ddcb' for the counter example\n",
    "        \n",
    "        hunk_lines_data = {}\n",
    "        \n",
    "        hunk_lines = hunk.content.split('\\n')[:-1]  # remove trailing empty element, because '\\n' is terminator\n",
    "        hunk_count_m = 0\n",
    "        hunk_count_p = 0\n",
    "        hunk_offsets_m = {}\n",
    "        hunk_offsets_p = {}\n",
    "        for i, line in enumerate(hunk_lines):\n",
    "            line_type = line[:1]\n",
    "            if line_type == '-':\n",
    "                hunk_offsets_m[i] = hunk_count_m\n",
    "                hunk_count_m += 1\n",
    "            elif line_type == '+':\n",
    "                hunk_offsets_p[i] = hunk_count_p\n",
    "                hunk_count_p += 1\n",
    "            elif line_type == ' ':\n",
    "                hunk_count_m += 1\n",
    "                hunk_count_p += 1\n",
    "            else:\n",
    "                pass\n",
    "\n",
    "        for hunk_line_no, line in enumerate(hunk_lines):\n",
    "            pm = line[:1]\n",
    "            hunk_lines_data[hunk_line_no] = {\n",
    "                'in_hunk_idx': hunk_line_no,\n",
    "                'pm': pm,\n",
    "                'diff_line': line,\n",
    "            }\n",
    "            if pm == '-':\n",
    "                if hunk_line_no in hunk_offsets_m:\n",
    "                    hunk_lines_data[hunk_line_no]['file_line'] = hunk.old_start + hunk_offsets_m[hunk_line_no]\n",
    "                else:\n",
    "                    # possibly mark with 'file_line' set to -1\n",
    "                    pass  # mark as N/A, should not happen\n",
    "            if pm == '+':\n",
    "                if hunk_line_no in hunk_offsets_p:\n",
    "                    hunk_lines_data[hunk_line_no]['file_line'] = hunk.new_start + hunk_offsets_p[hunk_line_no]\n",
    "                else:\n",
    "                    # possibly mark with 'file_line' set to -1\n",
    "                    pass  # mark as N/A, should not happen\n",
    "\n",
    "        for label, lines_list in hunk.lines_verified.items():\n",
    "            for hunk_line_no in lines_list:\n",
    "                hunk_lines_data[hunk_line_no]['lines_verified'] = label\n",
    "\n",
    "        # saving in \"wide\" format, witn N/A values\n",
    "        for user_name, user_data in hunk.lines_manual.items():\n",
    "            for label, lines_list in user_data.items():\n",
    "                for hunk_line_no in lines_list:\n",
    "                    hunk_lines_data[hunk_line_no][f'user_{user_name}'] = label\n",
    "\n",
    "        for lines_data in hunk_lines_data.values():\n",
    "            # with modern Python, we could use `.append(hunk_info | lines_data)`\n",
    "            smartshark_records.append({\n",
    "                **hunk_info,\n",
    "                **lines_data,\n",
    "            })\n",
    "\n",
    "    print(f\"{len(smartshark_hunk_ids)=} (present in 'hunk_labels.json')\")\n",
    "    print(f\"{len(hunks_extra)=} (outside 'smartshark_1_2' used in Herbold et al., and present in 'hunk_labels.json')\")\n",
    "    print(f\"{len(hunks_miss)=} (with 3 or less annotators/users)\")\n",
    "    print(f\"{len(hunks_multi_issue)=} (commits that link to multiple bugs)\")\n",
    "    print(f\"{len(smartshark_records)=} lines (results of parsing)\")\n",
    "\n",
    "else:\n",
    "    print(f\"skipping (use_mongodb=={use_mongodb})\")\n",
    "\n",
    "print(\"\")  # to separate printing of %%time output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "b1b839d5-cd89-443f-825a-d788687c3277",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'project': 'ant-ivy',\n",
       "  'repository_url': 'https://github.com/apache/ant-ivy',\n",
       "  'issue_id': 'IVY-1391',\n",
       "  'revision_hash': 'fdb75911387f77912c37dcee0af4a2f2e0977f3c',\n",
       "  'file': 'test/java/org/apache/ivy/ant/IvyPublishTest.java',\n",
       "  'hunk_id': '5a82f5f7912063217d88ef35',\n",
       "  'old_start': 249,\n",
       "  'old_lines': 0,\n",
       "  'new_start': 250,\n",
       "  'new_lines': 49,\n",
       "  'in_hunk_idx': 0,\n",
       "  'pm': '+',\n",
       "  'diff_line': '+    ',\n",
       "  'file_line': 250,\n",
       "  'lines_verified': 'test_doc_whitespace',\n",
       "  'user_aaghamohammadi': 'whitespace',\n",
       "  'user_KKC': 'test',\n",
       "  'user_uGAVaWPHVmMZKjBZ': 'test',\n",
       "  'user_LDlnwznmYPWdNVSA': 'whitespace'},\n",
       " {'project': 'ant-ivy',\n",
       "  'repository_url': 'https://github.com/apache/ant-ivy',\n",
       "  'issue_id': 'IVY-1391',\n",
       "  'revision_hash': 'fdb75911387f77912c37dcee0af4a2f2e0977f3c',\n",
       "  'file': 'test/java/org/apache/ivy/ant/IvyPublishTest.java',\n",
       "  'hunk_id': '5a82f5f7912063217d88ef35',\n",
       "  'old_start': 249,\n",
       "  'old_lines': 0,\n",
       "  'new_start': 250,\n",
       "  'new_lines': 49,\n",
       "  'in_hunk_idx': 1,\n",
       "  'pm': '+',\n",
       "  'diff_line': '+    public void testMergeParentWithoutLocationAttribute() throws IOException, ParseException {',\n",
       "  'file_line': 251,\n",
       "  'lines_verified': 'test',\n",
       "  'user_aaghamohammadi': 'test',\n",
       "  'user_KKC': 'test',\n",
       "  'user_uGAVaWPHVmMZKjBZ': 'test',\n",
       "  'user_LDlnwznmYPWdNVSA': 'bugfix'}]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_records[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "0793b911-8b71-4fa2-805b-8c651d8ec5e6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'project': 'commons-scxml',\n",
       "  'repository_url': 'https://github.com/apache/commons-scxml.git',\n",
       "  'issue_id': 'SCXML-14',\n",
       "  'revision_hash': '2bff4be4f58a70359154b9b5752074586ff3c852',\n",
       "  'file': 'src/test/java/org/apache/commons/scxml/SCXMLExecutorTest.java',\n",
       "  'hunk_id': '5bef252783ad2f56e82c95d1',\n",
       "  'old_start': 125,\n",
       "  'old_lines': 0,\n",
       "  'new_start': 128,\n",
       "  'new_lines': 21,\n",
       "  'in_hunk_idx': 19,\n",
       "  'pm': '+',\n",
       "  'diff_line': '+    }',\n",
       "  'file_line': 147,\n",
       "  'lines_verified': 'test',\n",
       "  'user_aaghamohammadi': 'test',\n",
       "  'user_melkor54248': 'test',\n",
       "  'user_NGKRdSgaxGAtmSqP': 'test',\n",
       "  'user_bossenti': 'test'},\n",
       " {'project': 'commons-scxml',\n",
       "  'repository_url': 'https://github.com/apache/commons-scxml.git',\n",
       "  'issue_id': 'SCXML-14',\n",
       "  'revision_hash': '2bff4be4f58a70359154b9b5752074586ff3c852',\n",
       "  'file': 'src/test/java/org/apache/commons/scxml/SCXMLExecutorTest.java',\n",
       "  'hunk_id': '5bef252783ad2f56e82c95d1',\n",
       "  'old_start': 125,\n",
       "  'old_lines': 0,\n",
       "  'new_start': 128,\n",
       "  'new_lines': 21,\n",
       "  'in_hunk_idx': 20,\n",
       "  'pm': '+',\n",
       "  'diff_line': '+',\n",
       "  'file_line': 148,\n",
       "  'lines_verified': 'test',\n",
       "  'user_aaghamohammadi': 'test',\n",
       "  'user_melkor54248': 'test',\n",
       "  'user_NGKRdSgaxGAtmSqP': 'test',\n",
       "  'user_bossenti': 'test'}]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_records[-2:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "c1808899-7642-4446-b509-98604b85cfff",
   "metadata": {},
   "outputs": [],
   "source": [
    "if use_mongodb:\n",
    "    # store to disk\n",
    "    with open('../../herbold_line_labels.json', 'w') as file:\n",
    "        file.write(json_util.dumps(smartshark_records))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b94c56fd-d094-44a2-8440-b32d5401c4fd",
   "metadata": {},
   "source": [
    "The `'herbold_line_labels.json'` file size is 174M.\n",
    "\n",
    "```console\n",
    "$ file herbold_line_labels.json\n",
    "herbold_line_labels.json: ASCII text, with very long lines (65536), with no line terminators\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "19e5a63a-4b38-4306-8756-c59c69e6c6a9",
   "metadata": {},
   "outputs": [],
   "source": [
    "if use_mongodb:\n",
    "    # create dataframe\n",
    "    smartshark_df = pd.DataFrame.from_records(smartshark_records)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "cbfc91ac-dec7-46c7-b411-9e06f9cad636",
   "metadata": {},
   "outputs": [],
   "source": [
    "if use_mongodb:\n",
    "    # and store to disk\n",
    "    # NOTE: using .to_parquet() of .to_feather() requires either the fastparquet or pyarrow library\n",
    "    # NOTE: using .to_hdf() requires pytables library\n",
    "    # NOTE: reading from pickle requires ABI compatibility\n",
    "    smartshark_df.to_csv('../../herbold_line_labels.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e6a2bc24-b859-4114-b163-37e6a7217c63",
   "metadata": {},
   "source": [
    "The `'herbold_line_labels.pickle'` file size would be 203M<br>\n",
    "The `'herbold_line_labels.csv'` file size is 104M"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "b43297e7-7ba7-4bdf-837a-be59094ea264",
   "metadata": {},
   "outputs": [],
   "source": [
    "# last line of this subsection\n",
    "disconnect()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "314c2042-6b38-4b98-8748-c47a2719aab7",
   "metadata": {},
   "source": [
    "## Load serialized SmartSHARK data from CSV"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5501010b-e3c5-427f-a060-885455199d70",
   "metadata": {},
   "source": [
    "If you skipped the extraction of data from the MongoDB, you can just load the data from the provided CSV file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "67c1fb83-6528-433e-8225-49acccb4b50c",
   "metadata": {},
   "outputs": [],
   "source": [
    "if not use_mongodb:\n",
    "    # without `low_memory=False` we get DtypeWarning about columns having mixed types\n",
    "    smartshark_df = pd.read_csv('../../herbold_line_labels.csv', low_memory=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "53ea5954-55fb-4e22-8e28-22301ead39db",
   "metadata": {},
   "source": [
    "## The structure of the SmartSHARK data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "5f788d5c-a1c2-4cba-bbc9-585d949b204d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['project', 'repository_url', 'issue_id', 'revision_hash', 'file',\n",
       "       'hunk_id', 'old_start', 'old_lines', 'new_start', 'new_lines',\n",
       "       'in_hunk_idx', 'pm', 'diff_line', 'file_line', 'lines_verified',\n",
       "       'user_aaghamohammadi', 'user_KKC', 'user_uGAVaWPHVmMZKjBZ',\n",
       "       'user_LDlnwznmYPWdNVSA', 'user_riruk', 'user_omalam', 'user_sherbold',\n",
       "       'user_vijaybw', 'user_evidencebp', 'user_atrautsch', 'user_davisjam',\n",
       "       'user_Jiong', 'user_kaka727', 'user_dvmarcilio', 'user_bledel',\n",
       "       'user_CUE', 'user_bossenti', 'user_nikoHu', 'user_simin',\n",
       "       'user_grodrig', 'user_JjGyDTFSqlCnXIzk', 'user_NGKRdSgaxGAtmSqP',\n",
       "       'user_melkor54248', 'user_TeeKea', 'user_rcolomo', 'user_perfwxc',\n",
       "       'user_aserebrenik', 'user_lancelot', 'user_ladybug', 'user_ealbrec',\n",
       "       'user_matin', 'user_ivanomalavolta', 'user_ppp23', 'user_psingh',\n",
       "       'user_Pyrrhon', 'user_LastButNotLeast', 'user_ctreude', 'user_azhenley',\n",
       "       'user_sulir', 'user_turtle', 'user_Badger', 'user_ethemutku',\n",
       "       'user_aarvQeMWXRSETYJu', 'user_ekrisza', 'user_danielstrueber',\n",
       "       'user_Ruuku', 'user_erbel', 'user_Vale', 'user_turhanb', 'user_joydeba',\n",
       "       'user_matt', 'user_deepblue', 'user_tAkmSjmapQPEZrVe', 'user_ap',\n",
       "       'user_Apollo', 'user_rNSKGGMSeMXtAEET', 'user_kentu713',\n",
       "       'user_vAvXBcvvrFJwUfHc', 'user_atuz', 'user_JG', 'user_yyh',\n",
       "       'user_LuCH', 'user_LeBron', 'user_Mohammad'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "b3f79fd1-b5a9-46aa-b493-b521c849f7e1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['project',\n",
       " 'repository_url',\n",
       " 'issue_id',\n",
       " 'revision_hash',\n",
       " 'file',\n",
       " 'hunk_id',\n",
       " 'old_start',\n",
       " 'old_lines',\n",
       " 'new_start',\n",
       " 'new_lines',\n",
       " 'in_hunk_idx',\n",
       " 'pm',\n",
       " 'diff_line',\n",
       " 'file_line',\n",
       " 'lines_verified']"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_col_user = [col for col in smartshark_df.columns if col.startswith('user_')]\n",
    "smartshark_col_rest = [col for col in smartshark_df.columns if not col.startswith('user_')]\n",
    "\n",
    "smartshark_col_rest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "40e16183-8d93-4461-8356-c0a42db0f3db",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(296348, 79)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "005f170b-434d-4bc8-8557-8c83dda493e7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "project            object\n",
       "repository_url     object\n",
       "issue_id           object\n",
       "revision_hash      object\n",
       "file               object\n",
       "hunk_id            object\n",
       "old_start           int64\n",
       "old_lines           int64\n",
       "new_start           int64\n",
       "new_lines           int64\n",
       "in_hunk_idx         int64\n",
       "pm                 object\n",
       "diff_line          object\n",
       "file_line         float64\n",
       "lines_verified     object\n",
       "dtype: object"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_col_rest].dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "05e760fd-cf41-4ed1-92bf-d40ec83aa8ff",
   "metadata": {},
   "source": [
    "### Filter out lines without 'file_line'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67480f1b-92f1-4442-9a47-e94af54b1f63",
   "metadata": {},
   "source": [
    "The fact that dtype for 'file_line' column is float64 and not int64 suggest that there are some N/A values there.\n",
    "\n",
    "Let's examine this case."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "a966b08b-a585-4018-a65a-96d9a157a580",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "np.int64(5536)"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df['file_line'].isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "2500ff3f-a80e-4dd0-93df-0b5e0b55f346",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>project</th>\n",
       "      <th>repository_url</th>\n",
       "      <th>issue_id</th>\n",
       "      <th>revision_hash</th>\n",
       "      <th>file</th>\n",
       "      <th>hunk_id</th>\n",
       "      <th>old_start</th>\n",
       "      <th>old_lines</th>\n",
       "      <th>new_start</th>\n",
       "      <th>new_lines</th>\n",
       "      <th>in_hunk_idx</th>\n",
       "      <th>pm</th>\n",
       "      <th>diff_line</th>\n",
       "      <th>file_line</th>\n",
       "      <th>lines_verified</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-716</td>\n",
       "      <td>66310f553103e515ab41aa573aaf2e4101efa635</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyRepositoryRepor...</td>\n",
       "      <td>5a82f5bd912063217a88a212</td>\n",
       "      <td>90</td>\n",
       "      <td>12</td>\n",
       "      <td>93</td>\n",
       "      <td>18</td>\n",
       "      <td>15</td>\n",
       "      <td></td>\n",
       "      <td>for (int i = 0; i &lt; mrids.length;...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>191</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-716</td>\n",
       "      <td>66310f553103e515ab41aa573aaf2e4101efa635</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyRepositoryRepor...</td>\n",
       "      <td>5a82f5bd912063217a88a212</td>\n",
       "      <td>90</td>\n",
       "      <td>12</td>\n",
       "      <td>93</td>\n",
       "      <td>18</td>\n",
       "      <td>22</td>\n",
       "      <td></td>\n",
       "      <td>}</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>224</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-1391</td>\n",
       "      <td>fdb75911387f77912c37dcee0af4a2f2e0977f3c</td>\n",
       "      <td>test/java/org/apache/ivy/ant/extends/child1/iv...</td>\n",
       "      <td>5a82f5f7912063217d88ef39</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>27</td>\n",
       "      <td>\\</td>\n",
       "      <td>\\ No newline at end of file</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>380</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-1206</td>\n",
       "      <td>84fd1dd099e6d44cdc5a2d6718197022f257dce9</td>\n",
       "      <td>test/java/org/apache/ivy/ant/ivy-extends-extra...</td>\n",
       "      <td>5a82f5ac912063217a889507</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>27</td>\n",
       "      <td>\\</td>\n",
       "      <td>\\ No newline at end of file</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>413</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-716</td>\n",
       "      <td>66310f553103e515ab41aa573aaf2e4101efa635</td>\n",
       "      <td>src/java/org/apache/ivy/core/install/InstallEn...</td>\n",
       "      <td>5a82f5bd912063217a88a216</td>\n",
       "      <td>106</td>\n",
       "      <td>6</td>\n",
       "      <td>106</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296032</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-21</td>\n",
       "      <td>db1e09c815847d5d11d063bb925af61e1784ac40</td>\n",
       "      <td>src/main/java/org/apache/commons/scxml/model/S...</td>\n",
       "      <td>5bef252783ad2f56ef2c9673</td>\n",
       "      <td>307</td>\n",
       "      <td>11</td>\n",
       "      <td>291</td>\n",
       "      <td>65</td>\n",
       "      <td>36</td>\n",
       "      <td></td>\n",
       "      <td>long wait = 0L;</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296050</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-21</td>\n",
       "      <td>db1e09c815847d5d11d063bb925af61e1784ac40</td>\n",
       "      <td>src/main/java/org/apache/commons/scxml/model/S...</td>\n",
       "      <td>5bef252783ad2f56ef2c9673</td>\n",
       "      <td>307</td>\n",
       "      <td>11</td>\n",
       "      <td>291</td>\n",
       "      <td>65</td>\n",
       "      <td>54</td>\n",
       "      <td></td>\n",
       "      <td>try {</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296053</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-21</td>\n",
       "      <td>db1e09c815847d5d11d063bb925af61e1784ac40</td>\n",
       "      <td>src/main/java/org/apache/commons/scxml/model/S...</td>\n",
       "      <td>5bef252783ad2f56ef2c9673</td>\n",
       "      <td>307</td>\n",
       "      <td>11</td>\n",
       "      <td>291</td>\n",
       "      <td>65</td>\n",
       "      <td>57</td>\n",
       "      <td></td>\n",
       "      <td>} catch (NumberFormatException nf...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296058</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-21</td>\n",
       "      <td>db1e09c815847d5d11d063bb925af61e1784ac40</td>\n",
       "      <td>src/main/java/org/apache/commons/scxml/model/S...</td>\n",
       "      <td>5bef252783ad2f56ef2c9673</td>\n",
       "      <td>307</td>\n",
       "      <td>11</td>\n",
       "      <td>291</td>\n",
       "      <td>65</td>\n",
       "      <td>62</td>\n",
       "      <td></td>\n",
       "      <td>}</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296061</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-21</td>\n",
       "      <td>db1e09c815847d5d11d063bb925af61e1784ac40</td>\n",
       "      <td>src/main/java/org/apache/commons/scxml/model/S...</td>\n",
       "      <td>5bef252783ad2f56ef2c9673</td>\n",
       "      <td>307</td>\n",
       "      <td>11</td>\n",
       "      <td>291</td>\n",
       "      <td>65</td>\n",
       "      <td>65</td>\n",
       "      <td></td>\n",
       "      <td>}</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5536 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              project                               repository_url  issue_id  \\\n",
       "184           ant-ivy            https://github.com/apache/ant-ivy   IVY-716   \n",
       "191           ant-ivy            https://github.com/apache/ant-ivy   IVY-716   \n",
       "224           ant-ivy            https://github.com/apache/ant-ivy  IVY-1391   \n",
       "380           ant-ivy            https://github.com/apache/ant-ivy  IVY-1206   \n",
       "413           ant-ivy            https://github.com/apache/ant-ivy   IVY-716   \n",
       "...               ...                                          ...       ...   \n",
       "296032  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-21   \n",
       "296050  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-21   \n",
       "296053  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-21   \n",
       "296058  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-21   \n",
       "296061  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-21   \n",
       "\n",
       "                                   revision_hash  \\\n",
       "184     66310f553103e515ab41aa573aaf2e4101efa635   \n",
       "191     66310f553103e515ab41aa573aaf2e4101efa635   \n",
       "224     fdb75911387f77912c37dcee0af4a2f2e0977f3c   \n",
       "380     84fd1dd099e6d44cdc5a2d6718197022f257dce9   \n",
       "413     66310f553103e515ab41aa573aaf2e4101efa635   \n",
       "...                                          ...   \n",
       "296032  db1e09c815847d5d11d063bb925af61e1784ac40   \n",
       "296050  db1e09c815847d5d11d063bb925af61e1784ac40   \n",
       "296053  db1e09c815847d5d11d063bb925af61e1784ac40   \n",
       "296058  db1e09c815847d5d11d063bb925af61e1784ac40   \n",
       "296061  db1e09c815847d5d11d063bb925af61e1784ac40   \n",
       "\n",
       "                                                     file  \\\n",
       "184     src/java/org/apache/ivy/ant/IvyRepositoryRepor...   \n",
       "191     src/java/org/apache/ivy/ant/IvyRepositoryRepor...   \n",
       "224     test/java/org/apache/ivy/ant/extends/child1/iv...   \n",
       "380     test/java/org/apache/ivy/ant/ivy-extends-extra...   \n",
       "413     src/java/org/apache/ivy/core/install/InstallEn...   \n",
       "...                                                   ...   \n",
       "296032  src/main/java/org/apache/commons/scxml/model/S...   \n",
       "296050  src/main/java/org/apache/commons/scxml/model/S...   \n",
       "296053  src/main/java/org/apache/commons/scxml/model/S...   \n",
       "296058  src/main/java/org/apache/commons/scxml/model/S...   \n",
       "296061  src/main/java/org/apache/commons/scxml/model/S...   \n",
       "\n",
       "                         hunk_id  old_start  old_lines  new_start  new_lines  \\\n",
       "184     5a82f5bd912063217a88a212         90         12         93         18   \n",
       "191     5a82f5bd912063217a88a212         90         12         93         18   \n",
       "224     5a82f5f7912063217d88ef39          0          0          1         27   \n",
       "380     5a82f5ac912063217a889507          0          0          1         27   \n",
       "413     5a82f5bd912063217a88a216        106          6        106          5   \n",
       "...                          ...        ...        ...        ...        ...   \n",
       "296032  5bef252783ad2f56ef2c9673        307         11        291         65   \n",
       "296050  5bef252783ad2f56ef2c9673        307         11        291         65   \n",
       "296053  5bef252783ad2f56ef2c9673        307         11        291         65   \n",
       "296058  5bef252783ad2f56ef2c9673        307         11        291         65   \n",
       "296061  5bef252783ad2f56ef2c9673        307         11        291         65   \n",
       "\n",
       "        in_hunk_idx pm                                          diff_line  \\\n",
       "184              15                  for (int i = 0; i < mrids.length;...   \n",
       "191              22                                                     }   \n",
       "224              27  \\                        \\ No newline at end of file   \n",
       "380              27  \\                        \\ No newline at end of file   \n",
       "413               2                                                         \n",
       "...             ... ..                                                ...   \n",
       "296032           36                                       long wait = 0L;   \n",
       "296050           54                                                 try {   \n",
       "296053           57                  } catch (NumberFormatException nf...   \n",
       "296058           62                                                     }   \n",
       "296061           65                                                     }   \n",
       "\n",
       "        file_line lines_verified  \n",
       "184           NaN            NaN  \n",
       "191           NaN            NaN  \n",
       "224           NaN            NaN  \n",
       "380           NaN            NaN  \n",
       "413           NaN            NaN  \n",
       "...           ...            ...  \n",
       "296032        NaN            NaN  \n",
       "296050        NaN            NaN  \n",
       "296053        NaN            NaN  \n",
       "296058        NaN            NaN  \n",
       "296061        NaN            NaN  \n",
       "\n",
       "[5536 rows x 15 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_df['file_line'].isna()][smartshark_col_rest]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "506d0e03-c816-49b3-8831-1ca60e8e8cc8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pm\n",
       "     5363\n",
       "\\     173\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_df['file_line'].isna()]['pm'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "f2dc6b67-76a9-41bb-913e-e871b130fa69",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pm\n",
       "+    222624\n",
       "-     68188\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_df['file_line'].notna()]['pm'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "27ddbb21-91c3-4ba1-b895-daf50af9335b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "184       NaN\n",
       "191       NaN\n",
       "224       NaN\n",
       "380       NaN\n",
       "413       NaN\n",
       "         ... \n",
       "296032    NaN\n",
       "296050    NaN\n",
       "296053    NaN\n",
       "296058    NaN\n",
       "296061    NaN\n",
       "Name: lines_verified, Length: 5536, dtype: object"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_df['file_line'].isna()]['lines_verified']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "f228d190-0d0c-48dc-a368-a6cf478cd86d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "lines_verified\n",
       "test                   115073\n",
       "bugfix                  73056\n",
       "documentation           40644\n",
       "None                    25888\n",
       "no_bugfix               13076\n",
       "whitespace              11935\n",
       "refactoring              5297\n",
       "test_doc_whitespace      4475\n",
       "unrelated                1361\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_df['file_line'].notna()]['lines_verified'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "503138f4-176f-4356-965b-10d46a6bfb3c",
   "metadata": {},
   "source": [
    "It turns out, that all cases without 'file_line' are non-changed lines that have been (accidentally?) caught.\n",
    "\n",
    "It also turns out that all those cases did not pass the 'lines_verified' test.\n",
    "\n",
    "So let's drop them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "72ac6a41-e419-41dd-9144-63245d930030",
   "metadata": {},
   "outputs": [],
   "source": [
    "smartshark_df_sel = smartshark_df[smartshark_df['file_line'].notna()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "9cb7943f-788a-4044-adba-2e5bc21b88a8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We are left with 290812/296348 = 0.9813192597891668 elements\n"
     ]
    }
   ],
   "source": [
    "print(f\"We are left with {smartshark_df_sel.shape[0]}/{smartshark_df.shape[0]} = \"\n",
    "      f\"{smartshark_df_sel.shape[0]/smartshark_df.shape[0]} elements\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "f621e401-dd0e-4fbc-b607-b53d260582e7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "np.int64(290812)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_mask_1 = smartshark_df['file_line'].notna()\n",
    "smartshark_mask_1.sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b3819a0-25ad-48b7-8999-93aeb4aac060",
   "metadata": {},
   "source": [
    "### Consensus in SmartSHARK"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "d161d883-a0fa-4e74-979a-9e5b5b88bcea",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['user_aaghamohammadi',\n",
       " 'user_KKC',\n",
       " 'user_uGAVaWPHVmMZKjBZ',\n",
       " 'user_LDlnwznmYPWdNVSA',\n",
       " 'user_riruk',\n",
       " 'user_omalam',\n",
       " 'user_sherbold',\n",
       " 'user_vijaybw',\n",
       " 'user_evidencebp',\n",
       " 'user_atrautsch',\n",
       " 'user_davisjam',\n",
       " 'user_Jiong',\n",
       " 'user_kaka727',\n",
       " 'user_dvmarcilio',\n",
       " 'user_bledel',\n",
       " 'user_CUE',\n",
       " 'user_bossenti',\n",
       " 'user_nikoHu',\n",
       " 'user_simin',\n",
       " 'user_grodrig',\n",
       " 'user_JjGyDTFSqlCnXIzk',\n",
       " 'user_NGKRdSgaxGAtmSqP',\n",
       " 'user_melkor54248',\n",
       " 'user_TeeKea',\n",
       " 'user_rcolomo',\n",
       " 'user_perfwxc',\n",
       " 'user_aserebrenik',\n",
       " 'user_lancelot',\n",
       " 'user_ladybug',\n",
       " 'user_ealbrec',\n",
       " 'user_matin',\n",
       " 'user_ivanomalavolta',\n",
       " 'user_ppp23',\n",
       " 'user_psingh',\n",
       " 'user_Pyrrhon',\n",
       " 'user_LastButNotLeast',\n",
       " 'user_ctreude',\n",
       " 'user_azhenley',\n",
       " 'user_sulir',\n",
       " 'user_turtle',\n",
       " 'user_Badger',\n",
       " 'user_ethemutku',\n",
       " 'user_aarvQeMWXRSETYJu',\n",
       " 'user_ekrisza',\n",
       " 'user_danielstrueber',\n",
       " 'user_Ruuku',\n",
       " 'user_erbel',\n",
       " 'user_Vale',\n",
       " 'user_turhanb',\n",
       " 'user_joydeba',\n",
       " 'user_matt',\n",
       " 'user_deepblue',\n",
       " 'user_tAkmSjmapQPEZrVe',\n",
       " 'user_ap',\n",
       " 'user_Apollo',\n",
       " 'user_rNSKGGMSeMXtAEET',\n",
       " 'user_kentu713',\n",
       " 'user_vAvXBcvvrFJwUfHc',\n",
       " 'user_atuz',\n",
       " 'user_JG',\n",
       " 'user_yyh',\n",
       " 'user_LuCH',\n",
       " 'user_LeBron',\n",
       " 'user_Mohammad']"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_col_user"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "259ac3df-4d26-4529-bf43-2edff7a19a33",
   "metadata": {},
   "outputs": [],
   "source": [
    "def most_common_value(row):\n",
    "    ret = Counter()\n",
    "    val = None\n",
    "    cnt = 0\n",
    "    for u_col in smartshark_col_user:\n",
    "        u_val = row[u_col]\n",
    "        if not pd.isnull(u_val):\n",
    "            ret[u_val] += 1\n",
    "            val = u_val\n",
    "            cnt += 1\n",
    "\n",
    "    if cnt > 0:\n",
    "        return Counter(ret).most_common(1)[0][0]\n",
    "    else:\n",
    "        return pd.NA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "24260c78-81d4-4835-932f-75b51b8d35ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "def n_reviewers(row):\n",
    "    ret = 0\n",
    "    for u_col in smartshark_col_user:\n",
    "        u_val = row[u_col]\n",
    "        if not pd.isnull(u_val):\n",
    "            ret += 1\n",
    "\n",
    "    return ret"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "eeea3405-22b6-4d3f-a03e-afed848e7e50",
   "metadata": {},
   "outputs": [],
   "source": [
    "def most_common_count(row):\n",
    "    ret = Counter()\n",
    "    val = None\n",
    "    cnt = 0\n",
    "    for u_col in smartshark_col_user:\n",
    "        u_val = row[u_col]\n",
    "        if not pd.isnull(u_val):\n",
    "            ret[u_val] += 1\n",
    "            val = u_val\n",
    "            cnt += 1\n",
    "\n",
    "    if cnt > 0:\n",
    "        return Counter(ret).most_common(1)[0][1]\n",
    "    else:\n",
    "        return np.nan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "985b8058-bdc1-43d8-8246-60fb68118cd6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "project                                                    ant-ivy\n",
       "repository_url                   https://github.com/apache/ant-ivy\n",
       "issue_id                                                  IVY-1391\n",
       "revision_hash             fdb75911387f77912c37dcee0af4a2f2e0977f3c\n",
       "file              test/java/org/apache/ivy/ant/IvyPublishTest.java\n",
       "                                        ...                       \n",
       "user_JG                                                        NaN\n",
       "user_yyh                                                       NaN\n",
       "user_LuCH                                                      NaN\n",
       "user_LeBron                                                    NaN\n",
       "user_Mohammad                                                  NaN\n",
       "Name: 0, Length: 79, dtype: object"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "row = smartshark_df.iloc[0]\n",
    "row"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "99773b57-e3a8-46d7-93c9-2aacae13ca63",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "user_aaghamohammadi      whitespace\n",
       "user_KKC                       test\n",
       "user_uGAVaWPHVmMZKjBZ          test\n",
       "user_LDlnwznmYPWdNVSA    whitespace\n",
       "Name: 0, dtype: object"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "row[smartshark_col_user].dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "5ea6a37b-e151-4932-bc37-c2a6b666769d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'whitespace'"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "most_common_value(row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "186ca88f-3a57-47f3-bd9c-e935bc94913f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_reviewers(row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "3004530a-1ca0-4001-b561-3d64bb5115bd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "most_common_count(row)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "550d8bd1-f7c0-4054-b7a3-782f3f86c4ae",
   "metadata": {},
   "source": [
    "Test for row without 'file_line'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "4836d1ca-c7c7-43df-b76f-fc096a7a8998",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "project                                               commons-scxml\n",
       "repository_url          https://github.com/apache/commons-scxml.git\n",
       "issue_id                                                   SCXML-21\n",
       "revision_hash              db1e09c815847d5d11d063bb925af61e1784ac40\n",
       "file              src/main/java/org/apache/commons/scxml/model/S...\n",
       "                                        ...                        \n",
       "user_JG                                                         NaN\n",
       "user_yyh                                                        NaN\n",
       "user_LuCH                                                       NaN\n",
       "user_LeBron                                                     NaN\n",
       "user_Mohammad                                                   NaN\n",
       "Name: 296061, Length: 79, dtype: object"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "row_na = smartshark_df.iloc[296061]\n",
    "row_na"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "fea3cec3-9410-45e6-b036-ad3ff65c21fa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series([], Name: 296061, dtype: object)"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "row_na[smartshark_col_user].dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "213c2e35-23df-4c54-9e47-4632157b57d0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<NA>"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "most_common_value(row_na)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "8f8eb57f-d9ff-4f0e-adbd-f3f1537aee66",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_reviewers(row_na)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "570c2efd-fe78-4066-8d3d-6a4a4f4ad9f5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "nan"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "most_common_count(row_na)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "23cd8ff8-6de5-41b7-ba71-e0c18fbba42c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<NA>"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.NA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "0d87f959-8f4e-4b76-aab1-9dc1037205c9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "nan"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.nan"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ccb69ddb-f464-40fc-8c9f-e10aa20b1d86",
   "metadata": {},
   "source": [
    "This code is not optimized, and takes around 1min 34s to execute."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "f867eb3e-4ae2-402c-a7b7-de30e8543f23",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 30s, sys: 350 ms, total: 1min 30s\n",
      "Wall time: 1min 30s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Index(['project', 'repository_url', 'issue_id', 'revision_hash', 'file',\n",
       "       'hunk_id', 'old_start', 'old_lines', 'new_start', 'new_lines',\n",
       "       'in_hunk_idx', 'pm', 'diff_line', 'file_line', 'lines_verified',\n",
       "       'user_aaghamohammadi', 'user_KKC', 'user_uGAVaWPHVmMZKjBZ',\n",
       "       'user_LDlnwznmYPWdNVSA', 'user_riruk', 'user_omalam', 'user_sherbold',\n",
       "       'user_vijaybw', 'user_evidencebp', 'user_atrautsch', 'user_davisjam',\n",
       "       'user_Jiong', 'user_kaka727', 'user_dvmarcilio', 'user_bledel',\n",
       "       'user_CUE', 'user_bossenti', 'user_nikoHu', 'user_simin',\n",
       "       'user_grodrig', 'user_JjGyDTFSqlCnXIzk', 'user_NGKRdSgaxGAtmSqP',\n",
       "       'user_melkor54248', 'user_TeeKea', 'user_rcolomo', 'user_perfwxc',\n",
       "       'user_aserebrenik', 'user_lancelot', 'user_ladybug', 'user_ealbrec',\n",
       "       'user_matin', 'user_ivanomalavolta', 'user_ppp23', 'user_psingh',\n",
       "       'user_Pyrrhon', 'user_LastButNotLeast', 'user_ctreude', 'user_azhenley',\n",
       "       'user_sulir', 'user_turtle', 'user_Badger', 'user_ethemutku',\n",
       "       'user_aarvQeMWXRSETYJu', 'user_ekrisza', 'user_danielstrueber',\n",
       "       'user_Ruuku', 'user_erbel', 'user_Vale', 'user_turhanb', 'user_joydeba',\n",
       "       'user_matt', 'user_deepblue', 'user_tAkmSjmapQPEZrVe', 'user_ap',\n",
       "       'user_Apollo', 'user_rNSKGGMSeMXtAEET', 'user_kentu713',\n",
       "       'user_vAvXBcvvrFJwUfHc', 'user_atuz', 'user_JG', 'user_yyh',\n",
       "       'user_LuCH', 'user_LeBron', 'user_Mohammad', 'most_common',\n",
       "       'n_reviewers', 'common_count'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "# this could be done in  single step, I think, by returning list or pd.Series\n",
    "\n",
    "smartshark_df['most_common'] = pd.DataFrame(smartshark_df.apply(most_common_value, axis=1))\n",
    "smartshark_df['n_reviewers'] = pd.DataFrame(smartshark_df.apply(n_reviewers, axis=1))\n",
    "smartshark_df['common_count'] = pd.DataFrame(smartshark_df.apply(most_common_count, axis=1))\n",
    "\n",
    "smartshark_df.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a391fc9d-acf6-48d5-948a-13a55bb6e8e9",
   "metadata": {},
   "source": [
    "From the Herbold et al. 2022 paper\n",
    "\n",
    "> Each commit was shown to four participants. Consensus is achieved if at least three participants agree on the same label. If this is not the case, no consensus for the line is achieved, i.e., the participants could not clearly identify which type of change a line is."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "37e291b5-3d96-47d1-ae22-00e2a56ef424",
   "metadata": {},
   "source": [
    "Analyze values in just generated columns: 'most_common', 'n_reviewers', and 'common_count'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "b80d03e9-b801-42be-ad8d-ea4eb04c9427",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "n_reviewers\n",
       "4     290572\n",
       "0       5543\n",
       "43       219\n",
       "64        14\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df['n_reviewers'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "3c8d42e3-65f0-43cc-b816-399e28de18dd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "common_count\n",
       "4.0     172544\n",
       "3.0      74589\n",
       "2.0      42046\n",
       "1.0       1393\n",
       "40.0        90\n",
       "39.0        40\n",
       "42.0        31\n",
       "41.0        31\n",
       "58.0         7\n",
       "38.0         5\n",
       "26.0         5\n",
       "36.0         4\n",
       "43.0         4\n",
       "57.0         3\n",
       "61.0         2\n",
       "31.0         2\n",
       "37.0         2\n",
       "54.0         1\n",
       "52.0         1\n",
       "33.0         1\n",
       "22.0         1\n",
       "32.0         1\n",
       "27.0         1\n",
       "30.0         1\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df['common_count'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "51f74193-380e-4ac8-88f1-b9370af0ee23",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common\n",
       "test             121370\n",
       "bugfix            82484\n",
       "documentation     49381\n",
       "whitespace        18552\n",
       "unrelated          9656\n",
       "refactoring        9362\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df['most_common'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ce16358-dea9-44f9-b65c-7125e7b2a43b",
   "metadata": {},
   "source": [
    "Compute consensus according to the definition in Herbold et al. 2022 paper: 3+ anotators agree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "add0805c-5f7b-4098-a295-56ec239a04b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "has_consensus\n",
       "True     247366\n",
       "False     48982\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df['has_consensus'] = smartshark_df['common_count'] >= 3\n",
    "smartshark_df['has_consensus'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "94ded5f5-9e86-425f-a037-0f9de8908e8f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "has_consensus\n",
       "True     0.834715\n",
       "False    0.165285\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df['has_consensus'].value_counts()/smartshark_df.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "708312fc-fa9e-4f74-9709-2e5b7d592ba8",
   "metadata": {},
   "source": [
    "Let's filter out elements without consensus (next filter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "08ad7af6-8b6f-468c-a53c-d8487147da5c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "np.int64(247366)"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_mask_2 = smartshark_df['has_consensus']\n",
    "smartshark_mask_2.sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "fe5b2e97-1eb0-4aba-b939-5fc1fc1d428f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "np.int64(247366)"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_mask = smartshark_mask_1 & smartshark_mask_2\n",
    "smartshark_mask.sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "fca5f846-195f-4b32-b35b-a72422f9fe00",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(247366, 83)"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df_sel = smartshark_df[smartshark_mask]\n",
    "smartshark_df_sel.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "9cf9123a-3074-48a1-9de8-aa5bda982dfa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We are left with 247366/296348 = 0.834714592303643 elements\n"
     ]
    }
   ],
   "source": [
    "print(f\"We are left with {smartshark_df_sel.shape[0]}/{smartshark_df.shape[0]} = \"\n",
    "      f\"{smartshark_df_sel.shape[0]/smartshark_df.shape[0]} elements\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "b415825f-b3e2-4ce7-ace9-0274e4f3f099",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['project',\n",
       " 'repository_url',\n",
       " 'issue_id',\n",
       " 'revision_hash',\n",
       " 'file',\n",
       " 'hunk_id',\n",
       " 'old_start',\n",
       " 'old_lines',\n",
       " 'new_start',\n",
       " 'new_lines',\n",
       " 'in_hunk_idx',\n",
       " 'pm',\n",
       " 'diff_line',\n",
       " 'file_line',\n",
       " 'lines_verified',\n",
       " 'most_common',\n",
       " 'n_reviewers',\n",
       " 'common_count',\n",
       " 'has_consensus']"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_col_user = [col for col in smartshark_df.columns if col.startswith('user_')]\n",
    "smartshark_col_rest = [col for col in smartshark_df.columns if not col.startswith('user_')]\n",
    "\n",
    "smartshark_col_rest"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8712a756-10a1-4358-8374-78215cf423b3",
   "metadata": {},
   "source": [
    "## DataFrame from diff-annotator on SmartSHARK repos"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33148dac-1b3b-4edd-9ed3-22130420e9a2",
   "metadata": {},
   "source": [
    "First step is to generate annotation data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "598e07a6-55c4-4f5c-8bbb-f14ca283505d",
   "metadata": {},
   "source": [
    "### Commits for/per repo"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d41d34bf-fad9-42dc-b813-9ce1590c915d",
   "metadata": {},
   "source": [
    "Find all the commit, and which repos they belong to"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "804e3854-3040-407a-923f-b49fefa51a70",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>project</th>\n",
       "      <th>repository_url</th>\n",
       "      <th>issue_id</th>\n",
       "      <th>revision_hash</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-1391</td>\n",
       "      <td>fdb75911387f77912c37dcee0af4a2f2e0977f3c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-1471</td>\n",
       "      <td>fe487b3336f8279d8b74eff05c5ebfe6e01281f0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-787</td>\n",
       "      <td>fe05cf44d3eb2f00fde5c75d60079d34743c9a48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-716</td>\n",
       "      <td>66310f553103e515ab41aa573aaf2e4101efa635</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-180</td>\n",
       "      <td>ff3804849e0badfe6fee032f39dacb577be94c22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295322</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-53</td>\n",
       "      <td>7e0c5e4c94a114c9fb39a861a735cdae51afaf57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295327</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-53</td>\n",
       "      <td>f64ac4cf3fda13a6b7b2712b3f7275b3902eec60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295745</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-54</td>\n",
       "      <td>92d8176168551c5da5f7ecdbb5607a6f01f90a11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295915</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-14</td>\n",
       "      <td>2bff4be4f58a70359154b9b5752074586ff3c852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295969</th>\n",
       "      <td>commons-scxml</td>\n",
       "      <td>https://github.com/apache/commons-scxml.git</td>\n",
       "      <td>SCXML-21</td>\n",
       "      <td>db1e09c815847d5d11d063bb925af61e1784ac40</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3519 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              project                               repository_url  issue_id  \\\n",
       "0             ant-ivy            https://github.com/apache/ant-ivy  IVY-1391   \n",
       "49            ant-ivy            https://github.com/apache/ant-ivy  IVY-1471   \n",
       "50            ant-ivy            https://github.com/apache/ant-ivy   IVY-787   \n",
       "52            ant-ivy            https://github.com/apache/ant-ivy   IVY-716   \n",
       "53            ant-ivy            https://github.com/apache/ant-ivy   IVY-180   \n",
       "...               ...                                          ...       ...   \n",
       "295322  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-53   \n",
       "295327  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-53   \n",
       "295745  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-54   \n",
       "295915  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-14   \n",
       "295969  commons-scxml  https://github.com/apache/commons-scxml.git  SCXML-21   \n",
       "\n",
       "                                   revision_hash  \n",
       "0       fdb75911387f77912c37dcee0af4a2f2e0977f3c  \n",
       "49      fe487b3336f8279d8b74eff05c5ebfe6e01281f0  \n",
       "50      fe05cf44d3eb2f00fde5c75d60079d34743c9a48  \n",
       "52      66310f553103e515ab41aa573aaf2e4101efa635  \n",
       "53      ff3804849e0badfe6fee032f39dacb577be94c22  \n",
       "...                                          ...  \n",
       "295322  7e0c5e4c94a114c9fb39a861a735cdae51afaf57  \n",
       "295327  f64ac4cf3fda13a6b7b2712b3f7275b3902eec60  \n",
       "295745  92d8176168551c5da5f7ecdbb5607a6f01f90a11  \n",
       "295915  2bff4be4f58a70359154b9b5752074586ff3c852  \n",
       "295969  db1e09c815847d5d11d063bb925af61e1784ac40  \n",
       "\n",
       "[3519 rows x 4 columns]"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_commits_df = smartshark_df[smartshark_mask_1][['project', 'repository_url', 'issue_id', 'revision_hash']].drop_duplicates()\n",
    "smartshark_commits_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "0181ecd7-ea97-41c4-90f5-1b8ce2af1696",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['ant-ivy', 'commons-math', 'opennlp', 'archiva', 'commons-bcel',\n",
       "       'commons-beanutils', 'commons-codec', 'commons-collections',\n",
       "       'commons-compress', 'commons-configuration', 'commons-dbcp',\n",
       "       'commons-digester', 'commons-io', 'commons-jcs', 'commons-lang',\n",
       "       'commons-net', 'commons-validator', 'commons-vfs', 'giraph',\n",
       "       'gora', 'parquet-mr', 'wss4j', 'deltaspike', 'systemml', 'jspwiki',\n",
       "       'eagle', 'santuario-java', 'commons-imaging', 'commons-scxml'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_projects = smartshark_commits_df['project'].unique()\n",
    "smartshark_projects"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "60533493-0106-4474-a730-fe4ba3eb4221",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "29"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(smartshark_projects)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "206b33d1-681d-4feb-a98e-c01947bd7f89",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['ant-ivy',\n",
       " 'archiva',\n",
       " 'commons-bcel',\n",
       " 'commons-beanutils',\n",
       " 'commons-codec',\n",
       " 'commons-collections',\n",
       " 'commons-compress',\n",
       " 'commons-configuration',\n",
       " 'commons-dbcp',\n",
       " 'commons-digester',\n",
       " 'commons-imaging',\n",
       " 'commons-io',\n",
       " 'commons-jcs',\n",
       " 'commons-lang',\n",
       " 'commons-math',\n",
       " 'commons-net',\n",
       " 'commons-scxml',\n",
       " 'commons-validator',\n",
       " 'commons-vfs',\n",
       " 'deltaspike',\n",
       " 'eagle',\n",
       " 'giraph',\n",
       " 'gora',\n",
       " 'jspwiki',\n",
       " 'opennlp',\n",
       " 'parquet-mr',\n",
       " 'santuario-java',\n",
       " 'systemml',\n",
       " 'wss4j']"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(smartshark_projects)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "695678a1-6c4a-4e89-a47e-171f05f88842",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "set()"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# this path is specific to the computer this notebook was run\n",
    "smartshark_repos_dir = '/mnt/data/msr/smartshark_repositories'\n",
    "\n",
    "dir_list = [path.name for path in Path(smartshark_repos_dir).iterdir() if path.is_dir()]\n",
    "set(smartshark_projects) - set(dir_list)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e6552c8-9f1b-470a-91f7-f2e87c4819b8",
   "metadata": {},
   "source": [
    "On previous runs, the following projects were missing:\n",
    "- commons-imaging: https://github.com/apache/commons-imaging.git\n",
    "- parquet-mr: https://github.com/apache/parquet-mr.git -> https://github.com/apache/parquet-java.git\n",
    "- santuario-java: https://github.com/apache/santuario-java.git (archived)\n",
    "- systemml: https://github.com/apache/systemml.git -> https://github.com/apache/systemds\n",
    "- wss4j: https://github.com/apache/wss4j.git -> 404 (repository **no longer exists** at that URL) -> https://gitbox.apache.org/repos/asf/ws-wss4j.git"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "bb87be0e-7172-4498-982b-b1a4b2ab8907",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ant-ivy                        has 548 commits\n",
      "commons-math                   has 391 commits\n",
      "opennlp                        has 150 commits\n",
      "archiva                        has   5 commits\n",
      "commons-bcel                   has  52 commits\n",
      "commons-beanutils              has  61 commits\n",
      "commons-codec                  has  58 commits\n",
      "commons-collections            has  92 commits\n",
      "commons-compress               has 204 commits\n",
      "commons-configuration          has 253 commits\n",
      "commons-dbcp                   has  89 commits\n",
      "commons-digester               has  26 commits\n",
      "commons-io                     has 116 commits\n",
      "commons-jcs                    has  73 commits\n",
      "commons-lang                   has 225 commits\n",
      "commons-net                    has 176 commits\n",
      "commons-validator              has  75 commits\n",
      "commons-vfs                    has 118 commits\n",
      "giraph                         has 146 commits\n",
      "gora                           has  98 commits\n",
      "parquet-mr                     has 119 commits\n",
      "wss4j                          has 245 commits\n",
      "deltaspike                     has   8 commits\n",
      "systemml                       has   6 commits\n",
      "jspwiki                        has   1 commits\n",
      "eagle                          has   2 commits\n",
      "santuario-java                 has  95 commits\n",
      "commons-imaging                has  20 commits\n",
      "commons-scxml                  has  67 commits\n"
     ]
    }
   ],
   "source": [
    "smartshark_projects_shas = {}\n",
    "for project in smartshark_projects:\n",
    "    print(f\"{project:30s} has \", end=\"\")\n",
    "\n",
    "    smartshark_projects_shas[project] = smartshark_commits_df[smartshark_commits_df['project'] == project]['revision_hash'].unique()\n",
    "    print(f\"{len(smartshark_projects_shas[project]):3d} commits\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "e8bc5cf8-9de0-4665-96d9-a9685be5e1f9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3519"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum([len(shas) for shas in smartshark_projects_shas.values()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "72286121-3455-4b54-9bf8-99f682a3c964",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3519, 4)"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_commits_df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95acc8a9-8b48-4dd0-8b9d-fd5ab5317e21",
   "metadata": {},
   "source": [
    "### Generate script for running diff-annotate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "486aa284-d575-4ca0-ad86-effc356628ea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common\n",
       "test             115073\n",
       "bugfix            73068\n",
       "documentation     40650\n",
       "whitespace        11937\n",
       "refactoring        5297\n",
       "unrelated          1341\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_df['has_consensus']]['most_common'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "5d5b7ac6-4032-47ae-9e54-404e27ff185c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common\n",
       "test             0.465193\n",
       "bugfix           0.295384\n",
       "documentation    0.164331\n",
       "whitespace       0.048256\n",
       "refactoring      0.021414\n",
       "unrelated        0.005421\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_df['has_consensus']]['most_common'].value_counts()/smartshark_df['has_consensus'].sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bd4c8f5b-8fed-48b8-8f69-9ed6c00ed5cc",
   "metadata": {},
   "source": [
    "We need to decide on how `diff-annotate` is to annotate lines, what possible line types we can assign.\n",
    "\n",
    "PatchScope does not implement refactoring detection - it is out of scope of the project (for now?).  This means that it would not assign the \"refactoring\" label (5297 = 2.14%).  Similarly, we would not be able to assign the \"unrelated\" label (1341 = 0.54%).\n",
    "\n",
    "Currently PatchScope cannot detect whitespace-only **_changes_** (11937 = 4.83%), as it currently does not do any matching between removed and added lines.  We can however assign \"whitespace\" label to whitespace-only _lines_... but we need to check if it improves agreement or not.\n",
    "\n",
    "The paper does not say whether documentation in test should be labelled \"documentation\" or \"test\".  We can try either way.\n",
    "\n",
    "Most probably comments in the code should be considered \"documentation\", and changes that are not comments to a file in programming language should be considered \"bugfix\"."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e39af87-f1f3-441f-bd88-86d3298aeb18",
   "metadata": {},
   "source": [
    "It turns out (see `'01-compare_annotations.ipynb'`) that we do not need to write special line callback code for SmartSHARK-compatibile labelling; adjusting the rules with command line arguments like `--purpose-to-annotation` and a bit of post-processing would be a better solution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "fe4f2725-e8a5-49f8-b39d-9d1735db0993",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ..."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f287f528-a3ff-4333-ab42-92e63cca5490",
   "metadata": {},
   "source": [
    "Next step is generate the script that would run `diff-annotate from-repo` on SmartSHARK repos for SmartSHARK commits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "60f9dfa9-6c69-450d-b76c-af4c13f6e485",
   "metadata": {},
   "outputs": [],
   "source": [
    "run_smartshark_script = '../../run_annotation_smartshark_repos.sh'\n",
    "smartshark_callback_file_relative = 'smartshark_line_callback.py'  # relative to script workdir\n",
    "smartshark_repos_dir = '/mnt/data/msr/smartshark_repositories'  # NOTE: this is value for this local computer (!!!)\n",
    "smartshark_annotations_dir = '/mnt/data/python-diff-annotator/example_annotations/smartshark'  # NOTE: configure this\n",
    "\n",
    "# creating part 1 of 2 of the script\n",
    "with open(run_smartshark_script, 'wt') as fp:\n",
    "    print('#!/usr/bin/sh', file=fp)\n",
    "    print('', file=fp)\n",
    "    print(f'CALLBACK_FILE=\"{smartshark_callback_file_relative}\"', file=fp)\n",
    "    print(f'REPOS_DIR=\"{smartshark_repos_dir}\"', file=fp)\n",
    "    print(f'ANNOTATIONS_DIR=\"{smartshark_annotations_dir}\"', file=fp)\n",
    "    print('', file=fp)\n",
    "    print('if [ ! -f \"$CALLBACK_FILE\" ]; then', file=fp)\n",
    "    print('    echo \"Could not find file $CALLBACK_FILE\"', file=fp)\n",
    "    print('    echo \"You are in directory $PWD\"', file=fp)\n",
    "    print('    echo \"Change directory to the top dir of this repo $(git rev-parse --show-toplevel 2>/dev/null)\"', file=fp)\n",
    "    print('    exit 1', file=fp)\n",
    "    print('fi', file=fp)\n",
    "    print('if [ ! -d \"$REPOS_DIR\" ]; then', file=fp)\n",
    "    print('    echo \"Could not find directory $REPOS_DIR\"', file=fp)\n",
    "    print('    exit 2', file=fp)\n",
    "    print('fi', file=fp)\n",
    "    print('', file=fp)\n",
    "    print('echo \"running annotations on SmartSHARK repos for SmartSHARK buggy commits, v2\"', file=fp)\n",
    "    print('echo \"trying to generate the same line annotations as used by SmartSHARK dataset\"', file=fp)\n",
    "    print('echo \"saving annotations to $ANNOTATIONS_DIR\"', file=fp)\n",
    "    print('', file=fp)\n",
    "\n",
    "Path(run_smartshark_script).chmod(0o755)  # 0755/-rwxr-xr-x"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "254034c6-3a96-4984-8f8d-e3a3607cdae4",
   "metadata": {},
   "source": [
    "One thing to consider is whether to use `--use-fanout` option, or not.  We opted to not use it, for easier annotation-parsing code."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "7b1639d9-b439-420d-9aa1-cf18542cf989",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ant-ivy             \targ_length <= 22831\n",
      "commons-math        \targ_length <= 16404\n",
      "opennlp             \targ_length <= 6513\n",
      "archiva             \targ_length <= 568\n",
      "commons-bcel        \targ_length <= 2505\n",
      "commons-beanutils   \targ_length <= 2884\n",
      "commons-codec       \targ_length <= 2753\n",
      "commons-collections \targ_length <= 4159\n",
      "commons-compress    \targ_length <= 8745\n",
      "commons-configuration\targ_length <= 10764\n",
      "commons-dbcp        \targ_length <= 4022\n",
      "commons-digester    \targ_length <= 1447\n",
      "commons-io          \targ_length <= 5125\n",
      "commons-jcs         \targ_length <= 3364\n",
      "commons-lang        \targ_length <= 9598\n",
      "commons-net         \targ_length <= 7587\n",
      "commons-validator   \targ_length <= 3458\n",
      "commons-vfs         \targ_length <= 5209\n",
      "giraph              \targ_length <= 6347\n",
      "gora                \targ_length <= 4375\n",
      "parquet-mr          \targ_length <= 5248\n",
      "wss4j               \targ_length <= 10404\n",
      "deltaspike          \targ_length <= 697\n",
      "systemml            \targ_length <= 611\n",
      "jspwiki             \targ_length <= 404\n",
      "eagle               \targ_length <= 441\n",
      "santuario-java      \targ_length <= 4272\n",
      "commons-imaging     \targ_length <= 1199\n",
      "commons-scxml       \targ_length <= 3122\n"
     ]
    }
   ],
   "source": [
    "# creating part 2 of 2 of the script\n",
    "for repo_name, repo_shas in smartshark_projects_shas.items():\n",
    "    print(f\"{repo_name:20s}\", end='')\n",
    "    cmd_str = ''.join([\n",
    "        'diff-annotate ',\n",
    "        '--ext-to-language=\".pom:Maven POM\" ',\n",
    "        '--ext-to-language=\".plist:XML Property List\" ',\n",
    "        '--ext-to-language=\".jar:Java Archive\" ',\n",
    "        '--pattern-to-purpose=\"*.jar:other\" ',\n",
    "        '--filename-to-language=RELEASE_NOTES:Text ',\n",
    "        # OLD VERSION\n",
    "        #f'--line-callback=\"$CALLBACK_FILE\" ',\n",
    "        # NEW VERSION\n",
    "        '--purpose-to-annotation=test ',\n",
    "        '--purpose-to-annotation=documentation ',\n",
    "        'from-repo ',\n",
    "        f'--output-dir=\"$ANNOTATIONS_DIR/{repo_name}\" ',\n",
    "        f'\"$REPOS_DIR/{repo_name}\" --no-walk=sorted {\" \".join(repo_shas)}',\n",
    "    ])\n",
    "    print(\"\\targ_length <=\", len(cmd_str))\n",
    "    \n",
    "    with open(run_smartshark_script, 'at') as fp:\n",
    "        print(f\"# {repo_name}\", file=fp)\n",
    "        print(cmd_str, file=fp)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5e77a6e-4398-4490-9d01-331e186aca58",
   "metadata": {},
   "source": [
    "### Read annotations, create DataFrame"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "63b29f00-a80f-466e-9402-bd4a13b91fa1",
   "metadata": {},
   "source": [
    "If needed, change directory to top directory of the PatchScope repository (this repository).\n",
    "\n",
    "Run the just generated script, `./run_annotation_smartshark_repos.sh`, or with\n",
    "```console\n",
    "/usr/bin/time --verbose --append --output=time.run_smartshark.log ./run_annotation_smartshark_repos.sh\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "dcebcad0-b811-4c7a-ba61-53c84bc0be9d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/mnt/data/python-diff-annotator/example_annotations/smartshark'"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# NOTE: this configuration is specific to the computer the notebook runs on (!)\n",
    "smartshark_annotations_dir = '/mnt/data/python-diff-annotator/example_annotations/smartshark'\n",
    "smartshark_annotations_dir"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "29146594-1f6d-4335-84b3-9c00cd4586cb",
   "metadata": {},
   "source": [
    "The cell below takes 1.61 s time to compute"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "7bb6ced1-1c34-4fe3-8e12-78c5ef0fad2a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ant-ivy                548 commits, 2766 changed files, 43348 changed lines\n",
      "archiva                  5 commits,   62 changed files,  1941 changed lines\n",
      "commons-bcel            52 commits,  185 changed files,  4472 changed lines\n",
      "commons-beanutils       61 commits,  131 changed files,  4746 changed lines\n",
      "commons-codec           58 commits,  167 changed files,  5145 changed lines\n",
      "commons-collections     92 commits,  283 changed files,  6412 changed lines\n",
      "commons-compress       204 commits,  523 changed files,  8806 changed lines\n",
      "commons-configuration  253 commits,  853 changed files, 22563 changed lines\n",
      "commons-dbcp            89 commits,  268 changed files,  6811 changed lines\n",
      "commons-digester        26 commits,   76 changed files,  1400 changed lines\n",
      "commons-imaging         20 commits,   58 changed files,  1024 changed lines\n",
      "commons-io             116 commits,  232 changed files,  5226 changed lines\n",
      "commons-jcs             73 commits,  451 changed files, 15640 changed lines\n",
      "commons-lang           225 commits,  495 changed files, 11443 changed lines\n",
      "commons-math           391 commits, 1396 changed files, 42796 changed lines\n",
      "commons-net            176 commits,  368 changed files,  6099 changed lines\n",
      "commons-scxml           67 commits,  307 changed files,  6066 changed lines\n",
      "commons-validator       75 commits,  176 changed files,  2615 changed lines\n",
      "commons-vfs            118 commits,  297 changed files,  5799 changed lines\n",
      "deltaspike               8 commits,   15 changed files,   155 changed lines\n",
      "eagle                    2 commits,    3 changed files,    50 changed lines\n",
      "giraph                 146 commits,  570 changed files, 15302 changed lines\n",
      "gora                    98 commits,  356 changed files, 10748 changed lines\n",
      "jspwiki                  1 commits,    2 changed files,   120 changed lines\n",
      "opennlp                150 commits,  405 changed files,  7610 changed lines\n",
      "parquet-mr             119 commits,  435 changed files, 15973 changed lines\n",
      "santuario-java           0 commits,    0 changed files,     0 changed lines\n",
      "systemml                 6 commits,    9 changed files,   225 changed lines\n",
      "wss4j                    0 commits,    0 changed files,     0 changed lines\n",
      "\n",
      "CPU times: user 1.48 s, sys: 92.2 ms, total: 1.58 s\n",
      "Wall time: 1.57 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "records_from_repos = []\n",
    "\n",
    "for subdir in sorted(Path(smartshark_annotations_dir).iterdir()):\n",
    "    print(f\"{subdir.name:22s}\", end='')\n",
    "    repo = subdir.name\n",
    "    \n",
    "    n_commits = 0\n",
    "    n_files = 0\n",
    "    n_lines = 0\n",
    "\n",
    "    for json_file in subdir.glob('*.json'):\n",
    "        sha = json_file.name.split('.', maxsplit=1)[0]\n",
    "        n_commits += 1\n",
    "\n",
    "        with open(json_file, mode='r') as json_fp:\n",
    "            json_data = json.load(json_fp)\n",
    "\n",
    "        for patched_file, file_data in json_data['changes'].items():\n",
    "            if patched_file == '/dev/null':\n",
    "                continue\n",
    "\n",
    "            n_files += 1\n",
    "\n",
    "            n_lines_in_file = 0\n",
    "            for pm in list(\"-+\"):\n",
    "                if pm in file_data:\n",
    "                    n_lines_in_file += len(file_data[pm])\n",
    "\n",
    "            hunk_max_line = Counter()\n",
    "            for pm in list(\"-+\"):\n",
    "                if pm not in file_data:\n",
    "                    continue\n",
    "\n",
    "                for line_data in file_data[pm]:\n",
    "                    hunk_max_line[line_data['hunk_idx']] = max(\n",
    "                        hunk_max_line[line_data['hunk_idx']],\n",
    "                        line_data['in_hunk_chg_idx']\n",
    "                    )\n",
    "            \n",
    "            for pm in list(\"-+\"):\n",
    "                if pm not in file_data:\n",
    "                    continue\n",
    "\n",
    "                for line_data in file_data[pm]:\n",
    "                    n_lines += 1\n",
    "\n",
    "                    # original generic adjustment of labeling\n",
    "                    annotation = 'bugfix' if line_data['type'] == 'code' else line_data['type'],\n",
    "                    \n",
    "                    # adjusted for better fit\n",
    "                    if line_data['type'] == 'code':\n",
    "                        annotation = 'bugfix'  # just a name for code change in bugfix\n",
    "                    elif line_data['type'] == 'data':\n",
    "                        # we lost information about whether it is in documentation directory, or test directory\n",
    "                        annotation = 'documentation'  # most common case\n",
    "                    else:\n",
    "                        annotation = line_data['type']\n",
    "                    \n",
    "                    records_from_repos.append({\n",
    "                        'project': repo,\n",
    "                        'sha': sha,\n",
    "                        'file': patched_file,\n",
    "                        'pm': pm,\n",
    "                        'line_id': line_data['id'],\n",
    "                        'file_line': line_data['file_line_no'],\n",
    "                        'hunk_idx': line_data['hunk_idx'],\n",
    "                        'in_hunk_idx': line_data['in_hunk_chg_idx'],\n",
    "                        'annotation': annotation,\n",
    "                        'n_lines_in_file': n_lines_in_file,\n",
    "                        'n_lines_in_hunk': hunk_max_line[line_data['hunk_idx']]+1,\n",
    "                        'purpose': line_data['purpose'],\n",
    "                    })\n",
    "\n",
    "    print(f\" {n_commits:3d} commits, {n_files:4d} changed files, {n_lines:5d} changed lines\")\n",
    "\n",
    "print(\"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "6f8f8cd9-c540-4a3f-ab33-1602a9a8cb19",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ant-ivy                548 commits\n",
      "archiva                  5 commits\n",
      "commons-bcel            52 commits\n",
      "commons-beanutils       61 commits\n",
      "commons-codec           58 commits\n",
      "commons-collections     92 commits\n",
      "commons-compress       204 commits\n",
      "commons-configuration  253 commits\n",
      "commons-dbcp            89 commits\n",
      "commons-digester        26 commits\n",
      "commons-imaging         20 commits\n",
      "commons-io             116 commits\n",
      "commons-jcs             73 commits\n",
      "commons-lang           225 commits\n",
      "commons-math           391 commits\n",
      "commons-net            176 commits\n",
      "commons-scxml           67 commits\n",
      "commons-validator       75 commits\n",
      "commons-vfs            118 commits\n",
      "deltaspike               8 commits\n",
      "eagle                    2 commits\n",
      "giraph                 146 commits\n",
      "gora                    98 commits\n",
      "jspwiki                  1 commits\n",
      "opennlp                150 commits\n",
      "parquet-mr             119 commits\n",
      "santuario-java          95 commits\n",
      "systemml                 6 commits\n",
      "wss4j                  245 commits\n"
     ]
    }
   ],
   "source": [
    "for project in sorted(smartshark_projects):\n",
    "    print(f\"{project:22s} {len(smartshark_projects_shas[project]):3d} commits\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "5757d887-23e6-4851-a0b2-f2dcbbb8cf0f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ant-ivy                548 commits smartshark, 548 commits from-repo\n",
      "archiva                  5 commits smartshark,   5 commits from-repo\n",
      "commons-bcel            52 commits smartshark,  52 commits from-repo\n",
      "commons-beanutils       61 commits smartshark,  61 commits from-repo\n",
      "commons-codec           58 commits smartshark,  58 commits from-repo\n",
      "commons-collections     92 commits smartshark,  92 commits from-repo\n",
      "commons-compress       204 commits smartshark, 204 commits from-repo\n",
      "commons-configuration  253 commits smartshark, 253 commits from-repo\n",
      "commons-dbcp            89 commits smartshark,  89 commits from-repo\n",
      "commons-digester        26 commits smartshark,  26 commits from-repo\n",
      "commons-imaging         20 commits smartshark,  20 commits from-repo\n",
      "commons-io             116 commits smartshark, 116 commits from-repo\n",
      "commons-jcs             73 commits smartshark,  73 commits from-repo\n",
      "commons-lang           225 commits smartshark, 225 commits from-repo\n",
      "commons-math           391 commits smartshark, 391 commits from-repo\n",
      "commons-net            176 commits smartshark, 176 commits from-repo\n",
      "commons-scxml           67 commits smartshark,  67 commits from-repo\n",
      "commons-validator       75 commits smartshark,  75 commits from-repo\n",
      "commons-vfs            118 commits smartshark, 118 commits from-repo\n",
      "deltaspike               8 commits smartshark,   8 commits from-repo\n",
      "eagle                    2 commits smartshark,   2 commits from-repo\n",
      "giraph                 146 commits smartshark, 146 commits from-repo\n",
      "gora                    98 commits smartshark,  98 commits from-repo\n",
      "jspwiki                  1 commits smartshark,   1 commits from-repo\n",
      "opennlp                150 commits smartshark, 150 commits from-repo\n",
      "parquet-mr             119 commits smartshark, 119 commits from-repo\n",
      "santuario-java          95 commits smartshark,   0 commits from-repo *\n",
      "systemml                 6 commits smartshark,   6 commits from-repo\n",
      "wss4j                  245 commits smartshark,   0 commits from-repo *\n"
     ]
    }
   ],
   "source": [
    "projects_to_drop = []\n",
    "\n",
    "for project in sorted(smartshark_projects):\n",
    "    n_commits_smartshark = len(smartshark_projects_shas[project])\n",
    "    print(f\"{project:22s} {n_commits_smartshark:3d} commits smartshark,\", end='')\n",
    "    if Path(smartshark_annotations_dir).joinpath(project).is_dir():\n",
    "        n_commits_from_repo = len(list(Path(smartshark_annotations_dir).joinpath(project).glob('*.json')))\n",
    "        print(f\" {n_commits_from_repo:3d} commits from-repo\", end='')\n",
    "        if n_commits_smartshark != n_commits_from_repo:\n",
    "            projects_to_drop.append(project)\n",
    "            print(\" *\")\n",
    "        else:\n",
    "            print(\"\")\n",
    "    else:\n",
    "        print(\"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "6c469db4-ad00-4854-80dd-dc467097c679",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['santuario-java', 'wss4j']"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(projects_to_drop)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "680a1d3d-29a0-40e9-8d7b-6a4bf5aa0fad",
   "metadata": {},
   "source": [
    "Check random commit from one of those repositories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "9768543f-21b2-4f9f-b2e1-ccaef60de6c4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>project</th>\n",
       "      <th>repository_url</th>\n",
       "      <th>issue_id</th>\n",
       "      <th>revision_hash</th>\n",
       "      <th>file</th>\n",
       "      <th>hunk_id</th>\n",
       "      <th>old_start</th>\n",
       "      <th>old_lines</th>\n",
       "      <th>new_start</th>\n",
       "      <th>new_lines</th>\n",
       "      <th>...</th>\n",
       "      <th>user_atuz</th>\n",
       "      <th>user_JG</th>\n",
       "      <th>user_yyh</th>\n",
       "      <th>user_LuCH</th>\n",
       "      <th>user_LeBron</th>\n",
       "      <th>user_Mohammad</th>\n",
       "      <th>most_common</th>\n",
       "      <th>n_reviewers</th>\n",
       "      <th>common_count</th>\n",
       "      <th>has_consensus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>280116</th>\n",
       "      <td>santuario-java</td>\n",
       "      <td>https://github.com/apache/santuario-java.git</td>\n",
       "      <td>SANTUARIO-499</td>\n",
       "      <td>88876f7a6736476c4f61a0bf100180c08576da7c</td>\n",
       "      <td>src/test/java/org/apache/xml/security/test/dom...</td>\n",
       "      <td>5caee5994dd2d957413f018f</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>75</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>test</td>\n",
       "      <td>4</td>\n",
       "      <td>3.0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 83 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               project                                repository_url  \\\n",
       "280116  santuario-java  https://github.com/apache/santuario-java.git   \n",
       "\n",
       "             issue_id                             revision_hash  \\\n",
       "280116  SANTUARIO-499  88876f7a6736476c4f61a0bf100180c08576da7c   \n",
       "\n",
       "                                                     file  \\\n",
       "280116  src/test/java/org/apache/xml/security/test/dom...   \n",
       "\n",
       "                         hunk_id  old_start  old_lines  new_start  new_lines  \\\n",
       "280116  5caee5994dd2d957413f018f          0          0          1         75   \n",
       "\n",
       "        ...  user_atuz user_JG user_yyh  user_LuCH user_LeBron user_Mohammad  \\\n",
       "280116  ...        NaN     NaN      NaN        NaN         NaN           NaN   \n",
       "\n",
       "       most_common n_reviewers common_count has_consensus  \n",
       "280116        test           4          3.0          True  \n",
       "\n",
       "[1 rows x 83 columns]"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_df['project'] == projects_to_drop[0]].head(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d89e05ad-2cea-4e50-9c12-da378c19afc5",
   "metadata": {},
   "source": [
    "Lets check https://github.com/apache/santuario-java/commit/88876f7a6736476c4f61a0bf100180c08576da7c\n",
    "\n",
    "GitHub gives the following warning:\n",
    "> This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0bd385f4-b945-4855-b062-884d132f3085",
   "metadata": {},
   "source": [
    "Let's examine how many commits, and how many lines it is"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "43d0436f-24be-42c5-a523-a14d7a5b32a7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[95, 245]"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[len(smartshark_projects_shas[project]) for project in projects_to_drop]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "442b519f-4ee5-4ee2-9684-e9b65c073bb1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "340"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum([len(smartshark_projects_shas[project]) for project in projects_to_drop])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "feb13b5f-6cce-43ab-8e67-5e52a5ca3570",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(36874, 83)"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "projects_to_drop_df = smartshark_df[\n",
    "    (smartshark_df['project'] == projects_to_drop[0]) |\n",
    "    (smartshark_df['project'] == projects_to_drop[1])\n",
    "]\n",
    "projects_to_drop_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "32c72752-a4a1-486b-993d-054d85054ba1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "340"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "projects_to_drop_df['revision_hash'].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "a9d8da5b-1014-4ac4-843d-d66b08d1d21d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Copied from 01-compare_annotations.ipynb\n",
    "#projects_to_drop = ['santuario-java', 'wss4j']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "2bdddc70-3fef-4600-9bb5-5b7dbfe33fe6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'project': 'ant-ivy',\n",
       "  'sha': '535c0c34334862d47ad339a99d0933e5eba76ae0',\n",
       "  'file': 'CHANGES.txt',\n",
       "  'pm': '+',\n",
       "  'line_id': 0,\n",
       "  'file_line': 1,\n",
       "  'hunk_idx': 0,\n",
       "  'in_hunk_idx': 0,\n",
       "  'annotation': 'documentation',\n",
       "  'n_lines_in_file': 2,\n",
       "  'n_lines_in_hunk': 2,\n",
       "  'purpose': 'documentation'},\n",
       " {'project': 'ant-ivy',\n",
       "  'sha': '535c0c34334862d47ad339a99d0933e5eba76ae0',\n",
       "  'file': 'CHANGES.txt',\n",
       "  'pm': '+',\n",
       "  'line_id': 1,\n",
       "  'file_line': 2,\n",
       "  'hunk_idx': 0,\n",
       "  'in_hunk_idx': 1,\n",
       "  'annotation': 'documentation',\n",
       "  'n_lines_in_file': 2,\n",
       "  'n_lines_in_hunk': 2,\n",
       "  'purpose': 'documentation'}]"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "records_from_repos[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "eee0f440-46e6-4302-8edc-d6a29495426b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>project</th>\n",
       "      <th>sha</th>\n",
       "      <th>file</th>\n",
       "      <th>pm</th>\n",
       "      <th>line_id</th>\n",
       "      <th>file_line</th>\n",
       "      <th>hunk_idx</th>\n",
       "      <th>in_hunk_idx</th>\n",
       "      <th>annotation</th>\n",
       "      <th>n_lines_in_file</th>\n",
       "      <th>n_lines_in_hunk</th>\n",
       "      <th>purpose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>535c0c34334862d47ad339a99d0933e5eba76ae0</td>\n",
       "      <td>CHANGES.txt</td>\n",
       "      <td>+</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>documentation</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>documentation</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>535c0c34334862d47ad339a99d0933e5eba76ae0</td>\n",
       "      <td>CHANGES.txt</td>\n",
       "      <td>+</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>documentation</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>documentation</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>535c0c34334862d47ad339a99d0933e5eba76ae0</td>\n",
       "      <td>src/java/fr/jayasoft/ivy/ModuleId.java</td>\n",
       "      <td>-</td>\n",
       "      <td>3</td>\n",
       "      <td>55</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>7</td>\n",
       "      <td>3</td>\n",
       "      <td>programming</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>535c0c34334862d47ad339a99d0933e5eba76ae0</td>\n",
       "      <td>src/java/fr/jayasoft/ivy/ModuleId.java</td>\n",
       "      <td>-</td>\n",
       "      <td>7</td>\n",
       "      <td>58</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>7</td>\n",
       "      <td>3</td>\n",
       "      <td>programming</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>535c0c34334862d47ad339a99d0933e5eba76ae0</td>\n",
       "      <td>src/java/fr/jayasoft/ivy/ModuleId.java</td>\n",
       "      <td>-</td>\n",
       "      <td>10</td>\n",
       "      <td>60</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>7</td>\n",
       "      <td>3</td>\n",
       "      <td>programming</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252530</th>\n",
       "      <td>systemml</td>\n",
       "      <td>ebdf770c872429a03ca125f2d69b25db59d0fa8c</td>\n",
       "      <td>src/main/java/org/apache/sysml/runtime/matrix/...</td>\n",
       "      <td>+</td>\n",
       "      <td>25</td>\n",
       "      <td>2691</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>25</td>\n",
       "      <td>8</td>\n",
       "      <td>programming</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252531</th>\n",
       "      <td>systemml</td>\n",
       "      <td>ebdf770c872429a03ca125f2d69b25db59d0fa8c</td>\n",
       "      <td>src/main/java/org/apache/sysml/runtime/matrix/...</td>\n",
       "      <td>+</td>\n",
       "      <td>4</td>\n",
       "      <td>2719</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>25</td>\n",
       "      <td>1</td>\n",
       "      <td>programming</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252532</th>\n",
       "      <td>systemml</td>\n",
       "      <td>ebdf770c872429a03ca125f2d69b25db59d0fa8c</td>\n",
       "      <td>src/main/java/org/apache/sysml/runtime/matrix/...</td>\n",
       "      <td>+</td>\n",
       "      <td>4</td>\n",
       "      <td>2733</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>25</td>\n",
       "      <td>3</td>\n",
       "      <td>programming</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252533</th>\n",
       "      <td>systemml</td>\n",
       "      <td>ebdf770c872429a03ca125f2d69b25db59d0fa8c</td>\n",
       "      <td>src/main/java/org/apache/sysml/runtime/matrix/...</td>\n",
       "      <td>+</td>\n",
       "      <td>6</td>\n",
       "      <td>2735</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>25</td>\n",
       "      <td>3</td>\n",
       "      <td>programming</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252534</th>\n",
       "      <td>systemml</td>\n",
       "      <td>ebdf770c872429a03ca125f2d69b25db59d0fa8c</td>\n",
       "      <td>src/main/java/org/apache/sysml/runtime/matrix/...</td>\n",
       "      <td>+</td>\n",
       "      <td>7</td>\n",
       "      <td>2736</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>25</td>\n",
       "      <td>3</td>\n",
       "      <td>programming</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>252535 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         project                                       sha  \\\n",
       "0        ant-ivy  535c0c34334862d47ad339a99d0933e5eba76ae0   \n",
       "1        ant-ivy  535c0c34334862d47ad339a99d0933e5eba76ae0   \n",
       "2        ant-ivy  535c0c34334862d47ad339a99d0933e5eba76ae0   \n",
       "3        ant-ivy  535c0c34334862d47ad339a99d0933e5eba76ae0   \n",
       "4        ant-ivy  535c0c34334862d47ad339a99d0933e5eba76ae0   \n",
       "...          ...                                       ...   \n",
       "252530  systemml  ebdf770c872429a03ca125f2d69b25db59d0fa8c   \n",
       "252531  systemml  ebdf770c872429a03ca125f2d69b25db59d0fa8c   \n",
       "252532  systemml  ebdf770c872429a03ca125f2d69b25db59d0fa8c   \n",
       "252533  systemml  ebdf770c872429a03ca125f2d69b25db59d0fa8c   \n",
       "252534  systemml  ebdf770c872429a03ca125f2d69b25db59d0fa8c   \n",
       "\n",
       "                                                     file pm  line_id  \\\n",
       "0                                             CHANGES.txt  +        0   \n",
       "1                                             CHANGES.txt  +        1   \n",
       "2                  src/java/fr/jayasoft/ivy/ModuleId.java  -        3   \n",
       "3                  src/java/fr/jayasoft/ivy/ModuleId.java  -        7   \n",
       "4                  src/java/fr/jayasoft/ivy/ModuleId.java  -       10   \n",
       "...                                                   ... ..      ...   \n",
       "252530  src/main/java/org/apache/sysml/runtime/matrix/...  +       25   \n",
       "252531  src/main/java/org/apache/sysml/runtime/matrix/...  +        4   \n",
       "252532  src/main/java/org/apache/sysml/runtime/matrix/...  +        4   \n",
       "252533  src/main/java/org/apache/sysml/runtime/matrix/...  +        6   \n",
       "252534  src/main/java/org/apache/sysml/runtime/matrix/...  +        7   \n",
       "\n",
       "        file_line  hunk_idx  in_hunk_idx     annotation  n_lines_in_file  \\\n",
       "0               1         0            0  documentation                2   \n",
       "1               2         0            1  documentation                2   \n",
       "2              55         1            0         bugfix                7   \n",
       "3              58         1            1         bugfix                7   \n",
       "4              60         1            2         bugfix                7   \n",
       "...           ...       ...          ...            ...              ...   \n",
       "252530       2691         2            7         bugfix               25   \n",
       "252531       2719         3            0         bugfix               25   \n",
       "252532       2733         4            0         bugfix               25   \n",
       "252533       2735         4            1         bugfix               25   \n",
       "252534       2736         4            2         bugfix               25   \n",
       "\n",
       "        n_lines_in_hunk        purpose  \n",
       "0                     2  documentation  \n",
       "1                     2  documentation  \n",
       "2                     3    programming  \n",
       "3                     3    programming  \n",
       "4                     3    programming  \n",
       "...                 ...            ...  \n",
       "252530                8    programming  \n",
       "252531                1    programming  \n",
       "252532                3    programming  \n",
       "252533                3    programming  \n",
       "252534                3    programming  \n",
       "\n",
       "[252535 rows x 12 columns]"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from_repos_df = pd.DataFrame.from_records(records_from_repos)\n",
    "from_repos_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "d624c8ce-b203-4a9f-b7b9-e868a4db3e10",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "smartshark_df.shape[0]=296348\n",
      "smartshark_df_sel.shape[0]=247366\n",
      "from_repos_df.shape[0]=252535\n"
     ]
    }
   ],
   "source": [
    "print(f\"{smartshark_df.shape[0]=}\")\n",
    "print(f\"{smartshark_df_sel.shape[0]=}\")\n",
    "print(f\"{from_repos_df.shape[0]=}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30e14882-016b-4fb9-be9a-27e0945450a0",
   "metadata": {},
   "source": [
    "Save dataframe, for future retrieval"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "9cd23fb9-3e9e-41e0-8b2d-1d169dd8ae45",
   "metadata": {},
   "outputs": [],
   "source": [
    "from_repos_df.to_csv('../../herbold_from_repos.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e8c94f6-85dc-4ed3-8609-c7932f180f1f",
   "metadata": {},
   "source": [
    "The `'herbold_from_repos.csv'` file size is 39M, vs 104M for `'herbold_line_labels.csv'`"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fcda0b90-82a9-4f27-9c55-bdbccc3a2101",
   "metadata": {},
   "source": [
    "## Compare SmartSHARK and diff-annotate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "1676e4b4-8e51-466a-8295-b073fa92ca6b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "smartshark_df.shape=(296348, 83), smartshark_df_sel.shape=(247366, 83)\n",
      "from_repos_df.shape=(252535, 12)\n"
     ]
    }
   ],
   "source": [
    "print(f\"{smartshark_df.shape=}, {smartshark_df_sel.shape=}\")\n",
    "print(f\"{from_repos_df.shape=}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "80236cc7-8ef6-4bc1-9378-d3ed10f82713",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['project',\n",
       " 'repository_url',\n",
       " 'issue_id',\n",
       " 'revision_hash',\n",
       " 'file',\n",
       " 'hunk_id',\n",
       " 'old_start',\n",
       " 'old_lines',\n",
       " 'new_start',\n",
       " 'new_lines',\n",
       " 'in_hunk_idx',\n",
       " 'pm',\n",
       " 'diff_line',\n",
       " 'file_line',\n",
       " 'lines_verified',\n",
       " 'most_common',\n",
       " 'n_reviewers',\n",
       " 'common_count',\n",
       " 'has_consensus']"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_col_rest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "3e231b88-2cae-4b51-8a1d-d48fa4b5af6e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['project',\n",
       " 'sha',\n",
       " 'file',\n",
       " 'pm',\n",
       " 'line_id',\n",
       " 'file_line',\n",
       " 'hunk_idx',\n",
       " 'in_hunk_idx',\n",
       " 'annotation',\n",
       " 'n_lines_in_file',\n",
       " 'n_lines_in_hunk',\n",
       " 'purpose']"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(from_repos_df.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "395197fc-ece3-4dc5-a26b-c5da5347df31",
   "metadata": {},
   "source": [
    "### Join/merge for comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "d8bc8568-c43e-457b-a92b-4579d0307d29",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>project</th>\n",
       "      <th>repository_url</th>\n",
       "      <th>issue_id</th>\n",
       "      <th>revision_hash</th>\n",
       "      <th>file</th>\n",
       "      <th>hunk_id</th>\n",
       "      <th>old_start</th>\n",
       "      <th>old_lines</th>\n",
       "      <th>new_start</th>\n",
       "      <th>new_lines</th>\n",
       "      <th>in_hunk_idx</th>\n",
       "      <th>pm</th>\n",
       "      <th>diff_line</th>\n",
       "      <th>file_line</th>\n",
       "      <th>lines_verified</th>\n",
       "      <th>most_common</th>\n",
       "      <th>n_reviewers</th>\n",
       "      <th>common_count</th>\n",
       "      <th>has_consensus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-1391</td>\n",
       "      <td>fdb75911387f77912c37dcee0af4a2f2e0977f3c</td>\n",
       "      <td>test/java/org/apache/ivy/ant/IvyPublishTest.java</td>\n",
       "      <td>5a82f5f7912063217d88ef35</td>\n",
       "      <td>249</td>\n",
       "      <td>0</td>\n",
       "      <td>250</td>\n",
       "      <td>49</td>\n",
       "      <td>0</td>\n",
       "      <td>+</td>\n",
       "      <td>+</td>\n",
       "      <td>250.0</td>\n",
       "      <td>test_doc_whitespace</td>\n",
       "      <td>whitespace</td>\n",
       "      <td>4</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   project                     repository_url  issue_id  \\\n",
       "0  ant-ivy  https://github.com/apache/ant-ivy  IVY-1391   \n",
       "\n",
       "                              revision_hash  \\\n",
       "0  fdb75911387f77912c37dcee0af4a2f2e0977f3c   \n",
       "\n",
       "                                               file                   hunk_id  \\\n",
       "0  test/java/org/apache/ivy/ant/IvyPublishTest.java  5a82f5f7912063217d88ef35   \n",
       "\n",
       "   old_start  old_lines  new_start  new_lines  in_hunk_idx pm diff_line  \\\n",
       "0        249          0        250         49            0  +     +       \n",
       "\n",
       "   file_line       lines_verified most_common  n_reviewers  common_count  \\\n",
       "0      250.0  test_doc_whitespace  whitespace            4           2.0   \n",
       "\n",
       "   has_consensus  \n",
       "0          False  "
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_col_rest].head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "c662a96c-2df1-42a4-83ef-d329a2cdf84f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>project</th>\n",
       "      <th>sha</th>\n",
       "      <th>file</th>\n",
       "      <th>pm</th>\n",
       "      <th>line_id</th>\n",
       "      <th>file_line</th>\n",
       "      <th>hunk_idx</th>\n",
       "      <th>in_hunk_idx</th>\n",
       "      <th>annotation</th>\n",
       "      <th>n_lines_in_file</th>\n",
       "      <th>n_lines_in_hunk</th>\n",
       "      <th>purpose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>535c0c34334862d47ad339a99d0933e5eba76ae0</td>\n",
       "      <td>CHANGES.txt</td>\n",
       "      <td>+</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>documentation</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>documentation</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   project                                       sha         file pm  line_id  \\\n",
       "0  ant-ivy  535c0c34334862d47ad339a99d0933e5eba76ae0  CHANGES.txt  +        0   \n",
       "\n",
       "   file_line  hunk_idx  in_hunk_idx     annotation  n_lines_in_file  \\\n",
       "0          1         0            0  documentation                2   \n",
       "\n",
       "   n_lines_in_hunk        purpose  \n",
       "0                2  documentation  "
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from_repos_df.head(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "28bd313f-09b7-4b18-b09c-d5a3776e110b",
   "metadata": {},
   "source": [
    "The 'revision_hash' in `smartshark_df` and 'sha' in `from_repos_df` refer to the same value, namely SHA-1 object identifier of the bug-fixing commit with changes.\n",
    "\n",
    "`pd.merge()` can deal with differently named columns thanks to `left_on` and `right_on` parameters (to be used instead of `on` parameter), but that leads to having both of those key columns in the final output; one can get the same information from `indicator` column as from pattern of NA values in `left_on` and `right_on` columns.\n",
    "\n",
    "Let's use the more specific name, 'revision_hash'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "4f45d0a2-8d35-4684-ad26-31df2ac7670c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>project</th>\n",
       "      <th>revision_hash</th>\n",
       "      <th>file</th>\n",
       "      <th>pm</th>\n",
       "      <th>line_id</th>\n",
       "      <th>file_line</th>\n",
       "      <th>hunk_idx</th>\n",
       "      <th>in_hunk_idx</th>\n",
       "      <th>annotation</th>\n",
       "      <th>n_lines_in_file</th>\n",
       "      <th>n_lines_in_hunk</th>\n",
       "      <th>purpose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>535c0c34334862d47ad339a99d0933e5eba76ae0</td>\n",
       "      <td>CHANGES.txt</td>\n",
       "      <td>+</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>documentation</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>documentation</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   project                             revision_hash         file pm  line_id  \\\n",
       "0  ant-ivy  535c0c34334862d47ad339a99d0933e5eba76ae0  CHANGES.txt  +        0   \n",
       "\n",
       "   file_line  hunk_idx  in_hunk_idx     annotation  n_lines_in_file  \\\n",
       "0          1         0            0  documentation                2   \n",
       "\n",
       "   n_lines_in_hunk        purpose  \n",
       "0                2  documentation  "
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from_repos_df = from_repos_df.rename(columns={'sha': 'revision_hash'})\n",
    "from_repos_df.head(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70752c8f-bb1e-4fe3-93a7-321f3f846731",
   "metadata": {},
   "source": [
    "The key columns to join those results on are"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "066c21e3-4e04-4c2b-aa11-b55b42670873",
   "metadata": {},
   "outputs": [],
   "source": [
    "key_cols = ['project', 'revision_hash', 'file', 'pm', 'file_line']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "7567202a-59d5-4d53-b9b8-d94b14955120",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "project            29\n",
       "revision_hash    3519\n",
       "file             5156\n",
       "pm                  4\n",
       "file_line        4070\n",
       "dtype: int64"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[key_cols].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "bf4785d8-0ca8-4767-bfe3-4d467565e00c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "project            29\n",
       "revision_hash    3519\n",
       "file             5156\n",
       "pm                  2\n",
       "file_line        4070\n",
       "dtype: int64"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df[smartshark_mask_1][key_cols].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "94c1c685-a5af-4d46-a5b0-8b722b7dee8c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "project            29\n",
       "revision_hash    3493\n",
       "file             4814\n",
       "pm                  2\n",
       "file_line        4035\n",
       "dtype: int64"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_df_sel[key_cols].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "124f8f13-56b6-4c56-af4c-dab146d3c73a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "project            27\n",
       "revision_hash    3179\n",
       "file             4770\n",
       "pm                  2\n",
       "file_line        3791\n",
       "dtype: int64"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from_repos_df[key_cols].nunique()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ff0955ed-9207-47d1-90ca-58af211b108f",
   "metadata": {},
   "source": [
    "Note that the above is not unique number of combinations.\n",
    "\n",
    "Only number of unique projects ('project') and unique revision hashes ('revision_hash') is correct with respect to number of changes (well, there could be duplicate values for revision_hash in different projects if project were forks of each other).\n",
    "\n",
    "The 'file' just shows how many different file names are there in changes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "76db39a6-055c-4ea3-b1ff-a193d76216b6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>project</th>\n",
       "      <th>repository_url</th>\n",
       "      <th>issue_id</th>\n",
       "      <th>revision_hash</th>\n",
       "      <th>file</th>\n",
       "      <th>hunk_id</th>\n",
       "      <th>old_start</th>\n",
       "      <th>old_lines</th>\n",
       "      <th>new_start</th>\n",
       "      <th>new_lines</th>\n",
       "      <th>...</th>\n",
       "      <th>common_count</th>\n",
       "      <th>has_consensus</th>\n",
       "      <th>line_id</th>\n",
       "      <th>hunk_idx</th>\n",
       "      <th>in_hunk_idx_repos</th>\n",
       "      <th>annotation</th>\n",
       "      <th>n_lines_in_file</th>\n",
       "      <th>n_lines_in_hunk</th>\n",
       "      <th>purpose</th>\n",
       "      <th>indicator_column</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-639</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>5a82f5ba912063217a889fbf</td>\n",
       "      <td>112.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>112.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-639</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>5a82f5ba912063217a889fbf</td>\n",
       "      <td>112.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>112.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>True</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-639</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>5a82f5ba912063217a889fc0</td>\n",
       "      <td>122.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>122.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-639</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>5a82f5ba912063217a889fc0</td>\n",
       "      <td>122.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>122.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>True</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-639</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>5a82f5ba912063217a889fc1</td>\n",
       "      <td>272.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>272.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 91 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   project                     repository_url issue_id  \\\n",
       "0  ant-ivy  https://github.com/apache/ant-ivy  IVY-639   \n",
       "1  ant-ivy  https://github.com/apache/ant-ivy  IVY-639   \n",
       "2  ant-ivy  https://github.com/apache/ant-ivy  IVY-639   \n",
       "3  ant-ivy  https://github.com/apache/ant-ivy  IVY-639   \n",
       "4  ant-ivy  https://github.com/apache/ant-ivy  IVY-639   \n",
       "\n",
       "                              revision_hash  \\\n",
       "0  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "1  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "2  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "3  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "4  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "\n",
       "                                              file                   hunk_id  \\\n",
       "0  src/java/org/apache/ivy/ant/IvyAntSettings.java  5a82f5ba912063217a889fbf   \n",
       "1  src/java/org/apache/ivy/ant/IvyAntSettings.java  5a82f5ba912063217a889fbf   \n",
       "2  src/java/org/apache/ivy/ant/IvyAntSettings.java  5a82f5ba912063217a889fc0   \n",
       "3  src/java/org/apache/ivy/ant/IvyAntSettings.java  5a82f5ba912063217a889fc0   \n",
       "4  src/java/org/apache/ivy/ant/IvyAntSettings.java  5a82f5ba912063217a889fc1   \n",
       "\n",
       "   old_start  old_lines  new_start  new_lines  ...  common_count  \\\n",
       "0      112.0        2.0      112.0        2.0  ...           2.0   \n",
       "1      112.0        2.0      112.0        2.0  ...           3.0   \n",
       "2      122.0        2.0      122.0        2.0  ...           2.0   \n",
       "3      122.0        2.0      122.0        2.0  ...           3.0   \n",
       "4      272.0        1.0      272.0        1.0  ...           2.0   \n",
       "\n",
       "  has_consensus line_id  hunk_idx in_hunk_idx_repos annotation  \\\n",
       "0         False     5.0       0.0               0.0     bugfix   \n",
       "1          True     6.0       0.0               1.0     bugfix   \n",
       "2         False     5.0       1.0               0.0     bugfix   \n",
       "3          True     6.0       1.0               1.0     bugfix   \n",
       "4         False     4.0       2.0               0.0     bugfix   \n",
       "\n",
       "  n_lines_in_file n_lines_in_hunk      purpose indicator_column  \n",
       "0            30.0             2.0  programming             both  \n",
       "1            30.0             2.0  programming             both  \n",
       "2            30.0             2.0  programming             both  \n",
       "3            30.0             2.0  programming             both  \n",
       "4            30.0             1.0  programming             both  \n",
       "\n",
       "[5 rows x 91 columns]"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# there is no much sense do the join that includes invalid lines (in SmartSHARK data)\n",
    "smartshark_vs_from_repos_df_merge = pd.merge(\n",
    "    smartshark_df[smartshark_mask_1], from_repos_df,\n",
    "    how='outer', on=key_cols,\n",
    "    indicator=\"indicator_column\", suffixes=(\"_shark\", \"_repos\"),\n",
    ")\n",
    "smartshark_vs_from_repos_df_merge.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "8c326d6d-4dbf-4ce8-8f44-1b1413e30864",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "both          237393\n",
       "left_only      53419\n",
       "right_only     15142\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indicator_value_counts_s = smartshark_vs_from_repos_df_merge['indicator_column'].value_counts()\n",
    "indicator_value_counts_s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "3c688cbc-f082-46c5-b3eb-a0d8240214e1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "both          0.816311\n",
       "left_only     0.183689\n",
       "right_only    0.052068\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indicator_value_counts_s/smartshark_df[smartshark_mask_1].shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "92c66aca-7e7c-47de-8ab5-866b994b09da",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "both          0.940040\n",
       "left_only     0.211531\n",
       "right_only    0.059960\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indicator_value_counts_s/from_repos_df.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "85e2148f-7815-4cfb-94dd-5eba99178102",
   "metadata": {},
   "source": [
    "Let's limit to only lines with consensus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "67951a12-5502-4a1c-8aef-163578cfb229",
   "metadata": {},
   "outputs": [],
   "source": [
    "#consensus_indicator_value_counts_s = smartshark_vs_from_repos_df_merge[\n",
    "#    smartshark_vs_from_repos_df_merge['has_consensus'].isna() |\n",
    "#    (smartshark_vs_from_repos_df_merge['has_consensus'] == True)\n",
    "#]['indicator_column'].value_counts()\n",
    "#consensus_indicator_value_counts_s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "f3578940-e24a-4e36-a858-7c1737454ca5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>project</th>\n",
       "      <th>repository_url</th>\n",
       "      <th>issue_id</th>\n",
       "      <th>revision_hash</th>\n",
       "      <th>file</th>\n",
       "      <th>hunk_id</th>\n",
       "      <th>old_start</th>\n",
       "      <th>old_lines</th>\n",
       "      <th>new_start</th>\n",
       "      <th>new_lines</th>\n",
       "      <th>...</th>\n",
       "      <th>common_count</th>\n",
       "      <th>has_consensus</th>\n",
       "      <th>line_id</th>\n",
       "      <th>hunk_idx</th>\n",
       "      <th>in_hunk_idx_repos</th>\n",
       "      <th>annotation</th>\n",
       "      <th>n_lines_in_file</th>\n",
       "      <th>n_lines_in_hunk</th>\n",
       "      <th>purpose</th>\n",
       "      <th>indicator_column</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>right_only</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-639</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>5a82f5ba912063217a889fbf</td>\n",
       "      <td>112.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>112.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>True</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>right_only</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>https://github.com/apache/ant-ivy</td>\n",
       "      <td>IVY-639</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>5a82f5ba912063217a889fc0</td>\n",
       "      <td>122.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>122.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>True</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ant-ivy</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</td>\n",
       "      <td>src/java/org/apache/ivy/ant/IvyAntSettings.java</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>bugfix</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>programming</td>\n",
       "      <td>right_only</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 91 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   project                     repository_url issue_id  \\\n",
       "0  ant-ivy                                NaN      NaN   \n",
       "1  ant-ivy  https://github.com/apache/ant-ivy  IVY-639   \n",
       "2  ant-ivy                                NaN      NaN   \n",
       "3  ant-ivy  https://github.com/apache/ant-ivy  IVY-639   \n",
       "4  ant-ivy                                NaN      NaN   \n",
       "\n",
       "                              revision_hash  \\\n",
       "0  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "1  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "2  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "3  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "4  001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   \n",
       "\n",
       "                                              file                   hunk_id  \\\n",
       "0  src/java/org/apache/ivy/ant/IvyAntSettings.java                       NaN   \n",
       "1  src/java/org/apache/ivy/ant/IvyAntSettings.java  5a82f5ba912063217a889fbf   \n",
       "2  src/java/org/apache/ivy/ant/IvyAntSettings.java                       NaN   \n",
       "3  src/java/org/apache/ivy/ant/IvyAntSettings.java  5a82f5ba912063217a889fc0   \n",
       "4  src/java/org/apache/ivy/ant/IvyAntSettings.java                       NaN   \n",
       "\n",
       "   old_start  old_lines  new_start  new_lines  ...  common_count  \\\n",
       "0        NaN        NaN        NaN        NaN  ...           NaN   \n",
       "1      112.0        2.0      112.0        2.0  ...           3.0   \n",
       "2        NaN        NaN        NaN        NaN  ...           NaN   \n",
       "3      122.0        2.0      122.0        2.0  ...           3.0   \n",
       "4        NaN        NaN        NaN        NaN  ...           NaN   \n",
       "\n",
       "  has_consensus line_id  hunk_idx in_hunk_idx_repos annotation  \\\n",
       "0           NaN     5.0       0.0               0.0     bugfix   \n",
       "1          True     6.0       0.0               1.0     bugfix   \n",
       "2           NaN     5.0       1.0               0.0     bugfix   \n",
       "3          True     6.0       1.0               1.0     bugfix   \n",
       "4           NaN     4.0       2.0               0.0     bugfix   \n",
       "\n",
       "  n_lines_in_file n_lines_in_hunk      purpose indicator_column  \n",
       "0            30.0             2.0  programming       right_only  \n",
       "1            30.0             2.0  programming             both  \n",
       "2            30.0             2.0  programming       right_only  \n",
       "3            30.0             2.0  programming             both  \n",
       "4            30.0             1.0  programming       right_only  \n",
       "\n",
       "[5 rows x 91 columns]"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_vs_from_repos_df_merge_sel = pd.merge(\n",
    "    smartshark_df_sel, from_repos_df,\n",
    "    how='outer', on=key_cols,\n",
    "    indicator=\"indicator_column\", suffixes=(\"_shark\", \"_repos\"),\n",
    ")\n",
    "smartshark_vs_from_repos_df_merge_sel.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "843d2e18-a2c6-44e7-8fc3-631e48be9444",
   "metadata": {},
   "source": [
    "### Analysis: matched lines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "id": "492d6d22-3d0f-494e-b6fa-8f031723f06f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "both          203727\n",
       "right_only     48808\n",
       "left_only      43639\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indicator_value_counts_s_sel = smartshark_vs_from_repos_df_merge_sel['indicator_column'].value_counts()\n",
    "indicator_value_counts_s_sel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "id": "8337217b-aa88-48e7-907d-8f0039aff0dd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "both          0.823585\n",
       "right_only    0.197311\n",
       "left_only     0.176415\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indicator_value_counts_s_sel/smartshark_df_sel.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "id": "df10156c-97f9-4951-95db-ab87ff043ab0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "both          0.806728\n",
       "right_only    0.193272\n",
       "left_only     0.172804\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indicator_value_counts_s_sel/from_repos_df.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "id": "16324157-190d-411a-9e26-02403c80c086",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "project            27\n",
       "revision_hash    3179\n",
       "file             4494\n",
       "pm                  2\n",
       "file_line        3787\n",
       "dtype: int64"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_vs_from_repos_df_merge[\n",
    "    smartshark_vs_from_repos_df_merge['indicator_column'] == 'both'\n",
    "][key_cols].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "f0258766-fcc4-4dd4-8a02-ba9f4f817f6b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "project            27\n",
       "revision_hash    3153\n",
       "file             4178\n",
       "pm                  2\n",
       "file_line        3740\n",
       "dtype: int64"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_vs_from_repos_df_merge_sel[\n",
    "    smartshark_vs_from_repos_df_merge_sel['indicator_column'] == 'both'\n",
    "][key_cols].nunique()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0af1dc6f-2b5a-49a1-8b0f-3175c48c804f",
   "metadata": {},
   "source": [
    "Let's take for account problems with running PatchScope's `diff-annotate` on 2 out of 29 repositories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "e2a6322e-ebac-4a2c-873e-acbdf5e3f409",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['santuario-java', 'wss4j']"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "projects_to_drop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "id": "30399f35-20d9-4838-b92c-a7f2ee2a86e1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(269668, 91)"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_vs_from_repos_df_merge_x = smartshark_vs_from_repos_df_merge[\n",
    "    (smartshark_vs_from_repos_df_merge['project'] != 'santuario-java') &\n",
    "    (smartshark_vs_from_repos_df_merge['project'] != 'wss4j')\n",
    "]\n",
    "smartshark_vs_from_repos_df_merge_x.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "b2b5f6da-a31c-4deb-af4d-70e7248292f7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "both          237393\n",
       "left_only      17133\n",
       "right_only     15142\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indicator_value_counts_s_x = smartshark_vs_from_repos_df_merge_x['indicator_column'].value_counts()\n",
    "indicator_value_counts_s_x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "9f0d7569-1c57-4787-bae1-69297964982b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "both          0.914901\n",
       "left_only     0.066030\n",
       "right_only    0.058357\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indicator_value_counts_s_x/smartshark_df[\n",
    "    (smartshark_df['project'] != 'santuario-java') &\n",
    "    (smartshark_df['project'] != 'wss4j')\n",
    "].shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "b094bbd5-c6df-44be-96c9-41be1ff121fb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "both          0.940040\n",
       "left_only     0.067844\n",
       "right_only    0.059960\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indicator_value_counts_s_x/from_repos_df.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce0df60f-587c-45eb-ad8f-c7d19c141d02",
   "metadata": {},
   "source": [
    "### Analysis: annotations agreement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "id": "9e7ddda5-e59d-41e6-bb48-da5ea879cb14",
   "metadata": {},
   "outputs": [],
   "source": [
    "smartshark_vs_from_repos_df_merge['eq'] = \\\n",
    "    smartshark_vs_from_repos_df_merge['most_common'] == smartshark_vs_from_repos_df_merge['annotation']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "id": "d00edb21-ffb7-4e06-9757-b7fb182eebab",
   "metadata": {},
   "outputs": [],
   "source": [
    "smartshark_vs_from_repos_df_merge_sel['eq'] = \\\n",
    "    smartshark_vs_from_repos_df_merge_sel['most_common'] == smartshark_vs_from_repos_df_merge_sel['annotation']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "a649dfac-a25e-405d-9baf-f0b8b83d0a19",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "True     237393\n",
       "False     68561\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches_mask = (smartshark_vs_from_repos_df_merge['indicator_column'] == 'both')\n",
    "matches_mask.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "ce71a0b6-96a7-42f3-a40d-45be89d759c0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "indicator_column\n",
       "True     203727\n",
       "False     92447\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches_mask_sel = (smartshark_vs_from_repos_df_merge_sel['indicator_column'] == 'both')\n",
    "matches_mask_sel.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "44e252e8-9df6-4b84-bf9b-dbaf74b3c006",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "eq\n",
       "True     198252\n",
       "False     39141\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eq_s = smartshark_vs_from_repos_df_merge[matches_mask]['eq'].value_counts()\n",
    "eq_s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "id": "ada211c2-63b2-4eb1-9322-fb3e12cb9c5b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "eq\n",
       "True     185947\n",
       "False     17780\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eq_s_sel = smartshark_vs_from_repos_df_merge_sel[matches_mask_sel]['eq'].value_counts()\n",
    "eq_s_sel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "id": "9e14c43a-f03b-4ce9-b1bc-4ff1c4a350d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "eq\n",
       "True     0.912726\n",
       "False    0.087274\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eq_s_sel/smartshark_vs_from_repos_df_merge_sel[matches_mask_sel].shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6edbc74a-848d-40d2-860b-725cc24879a1",
   "metadata": {},
   "source": [
    "### Disagreement analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84a885ae-e5f1-445d-9044-fe1cc4313407",
   "metadata": {},
   "source": [
    "Let's create DataFrame for those lines for which there is match, and which have consensus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "3f72ee50-578c-47a2-97f7-d0b9b29b3018",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(203727, 92)"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_vs_from_repos_df_merge_match_sel = smartshark_vs_from_repos_df_merge_sel[matches_mask_sel]\n",
    "smartshark_vs_from_repos_df_merge_match_sel.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "1b2a84e9-73c9-4315-bfd5-62e421502162",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(237393, 92)"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# for comparison\n",
    "smartshark_vs_from_repos_df_merge[matches_mask].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "9ea4380e-6886-478b-8a0b-734c2a2cfe9b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common    annotation   \n",
       "test           test             92762\n",
       "bugfix         bugfix           59278\n",
       "documentation  documentation    33907\n",
       "whitespace     bugfix            9675\n",
       "refactoring    bugfix            2764\n",
       "documentation  test               870\n",
       "bugfix         documentation      824\n",
       "               test               822\n",
       "unrelated      bugfix             773\n",
       "test           bugfix             386\n",
       "whitespace     documentation      285\n",
       "unrelated      other              202\n",
       "documentation  markup             142\n",
       "refactoring    documentation      142\n",
       "unrelated      project            137\n",
       "documentation  project            136\n",
       "whitespace     test               122\n",
       "test           documentation      118\n",
       "documentation  bugfix              89\n",
       "unrelated      test                84\n",
       "bugfix         other               78\n",
       "               project             53\n",
       "refactoring    test                20\n",
       "unrelated      documentation       20\n",
       "whitespace     project             10\n",
       "               other                8\n",
       "test           other                8\n",
       "               project              7\n",
       "documentation  other                5\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smartshark_vs_from_repos_df_merge_match_sel[['most_common', 'annotation']].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "id": "0188736b-2c31-49df-a01f-b213c661ab6b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common    annotation   \n",
       "whitespace     bugfix           9675\n",
       "refactoring    bugfix           2764\n",
       "documentation  test              870\n",
       "bugfix         documentation     824\n",
       "               test              822\n",
       "unrelated      bugfix            773\n",
       "test           bugfix            386\n",
       "whitespace     documentation     285\n",
       "unrelated      other             202\n",
       "documentation  markup            142\n",
       "refactoring    documentation     142\n",
       "unrelated      project           137\n",
       "documentation  project           136\n",
       "whitespace     test              122\n",
       "test           documentation     118\n",
       "documentation  bugfix             89\n",
       "unrelated      test               84\n",
       "bugfix         other              78\n",
       "               project            53\n",
       "unrelated      documentation      20\n",
       "refactoring    test               20\n",
       "whitespace     project            10\n",
       "               other               8\n",
       "test           other               8\n",
       "               project             7\n",
       "documentation  other               5\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 130,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmp_counts_s = smartshark_vs_from_repos_df_merge_match_sel[\n",
    "    ~smartshark_vs_from_repos_df_merge_match_sel['eq'].astype(bool)\n",
    "][[\n",
    "    'most_common', 'annotation'\n",
    "]].value_counts()\n",
    "cmp_counts_s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "id": "c95b19c3-370d-4dc4-87f7-7c2026d7e917",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common    annotation   \n",
       "whitespace     bugfix           0.047490\n",
       "refactoring    bugfix           0.013567\n",
       "documentation  test             0.004270\n",
       "bugfix         documentation    0.004045\n",
       "               test             0.004035\n",
       "unrelated      bugfix           0.003794\n",
       "test           bugfix           0.001895\n",
       "whitespace     documentation    0.001399\n",
       "unrelated      other            0.000992\n",
       "documentation  markup           0.000697\n",
       "refactoring    documentation    0.000697\n",
       "unrelated      project          0.000672\n",
       "documentation  project          0.000668\n",
       "whitespace     test             0.000599\n",
       "test           documentation    0.000579\n",
       "documentation  bugfix           0.000437\n",
       "unrelated      test             0.000412\n",
       "bugfix         other            0.000383\n",
       "               project          0.000260\n",
       "unrelated      documentation    0.000098\n",
       "refactoring    test             0.000098\n",
       "whitespace     project          0.000049\n",
       "               other            0.000039\n",
       "test           other            0.000039\n",
       "               project          0.000034\n",
       "documentation  other            0.000025\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmp_counts_s/smartshark_vs_from_repos_df_merge_match_sel.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "id": "8d3e88c5-b7ee-4487-9891-9af20c4e2bdd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "eq\n",
       "False    185947\n",
       "True      17780\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "neq_mask = ~smartshark_vs_from_repos_df_merge_match_sel['eq'].astype(bool)\n",
    "neq_mask.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "id": "d21c24ee-29c5-4fce-9b80-ee0b06276c63",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common    annotation   \n",
       "whitespace     bugfix           0.544151\n",
       "refactoring    bugfix           0.155456\n",
       "documentation  test             0.048931\n",
       "bugfix         documentation    0.046344\n",
       "               test             0.046232\n",
       "unrelated      bugfix           0.043476\n",
       "test           bugfix           0.021710\n",
       "whitespace     documentation    0.016029\n",
       "unrelated      other            0.011361\n",
       "documentation  markup           0.007987\n",
       "refactoring    documentation    0.007987\n",
       "unrelated      project          0.007705\n",
       "documentation  project          0.007649\n",
       "whitespace     test             0.006862\n",
       "test           documentation    0.006637\n",
       "documentation  bugfix           0.005006\n",
       "unrelated      test             0.004724\n",
       "bugfix         other            0.004387\n",
       "               project          0.002981\n",
       "unrelated      documentation    0.001125\n",
       "refactoring    test             0.001125\n",
       "whitespace     project          0.000562\n",
       "               other            0.000450\n",
       "test           other            0.000450\n",
       "               project          0.000394\n",
       "documentation  other            0.000281\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmp_counts_s/neq_mask.value_counts()[True]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "id": "cc18aef8-55e0-41c9-bf4b-aa41ffb230b2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common  \n",
       "whitespace       10100\n",
       "refactoring       2926\n",
       "bugfix            1777\n",
       "documentation     1242\n",
       "unrelated         1216\n",
       "test               519\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmp_counts_smartshark_s = smartshark_vs_from_repos_df_merge_match_sel[\n",
    "    neq_mask\n",
    "][[\n",
    "    'most_common'\n",
    "]].value_counts()\n",
    "cmp_counts_smartshark_s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "id": "482593ac-7cad-40c4-b631-04e4a72ad490",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common  \n",
       "whitespace       0.049576\n",
       "refactoring      0.014362\n",
       "bugfix           0.008722\n",
       "documentation    0.006096\n",
       "unrelated        0.005969\n",
       "test             0.002548\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmp_counts_smartshark_s/smartshark_vs_from_repos_df_merge_match_sel.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "id": "fe4ce427-444c-46a3-bd88-2907d04f2c91",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "most_common  \n",
       "whitespace       0.568054\n",
       "refactoring      0.164567\n",
       "bugfix           0.099944\n",
       "documentation    0.069854\n",
       "unrelated        0.068391\n",
       "test             0.029190\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# of non matches\n",
    "cmp_counts_smartshark_s/neq_mask.value_counts()[True]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6bade024-200b-41a9-bfe5-f21910cc77b7",
   "metadata": {},
   "source": [
    "### Per commit agreement percentage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "id": "4507f2ce-9ea3-48d5-9d70-695b5b7fd9f1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "#from matplotlib.dates import MONTHLY, WEEKLY, DateFormatter, rrulewrapper, RRuleLocator\n",
    "#from pandas.plotting import register_matplotlib_converters\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "#register_matplotlib_converters()\n",
    "plt.rcParams.update({'font.size': 12})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "id": "535cede8-31f2-496e-b123-4567b2838b36",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sum</th>\n",
       "      <th>count</th>\n",
       "      <th>ratio</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>revision_hash</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>000e726af8464d94e7d02d307dde0e07e176d58c</th>\n",
       "      <td>2</td>\n",
       "      <td>18</td>\n",
       "      <td>0.111111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>001e2caa07a1dbfd92fbf86a26f30aaef23be5f5</th>\n",
       "      <td>51</td>\n",
       "      <td>66</td>\n",
       "      <td>0.772727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>002d7bf508ee59f23a04a9f21d02e9673dd38a71</th>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>002f8ec7aa01956b282ffd78062ee6fed6ab1eb9</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>004124ac5dbf5edbf925078652526267468821e7</th>\n",
       "      <td>165</td>\n",
       "      <td>166</td>\n",
       "      <td>0.993976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ffc38b1f3b35ddfadb942240b9d2f933ec83d1be</th>\n",
       "      <td>62</td>\n",
       "      <td>63</td>\n",
       "      <td>0.984127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ffc6adf7d493def070e26592bf828accbeecef2d</th>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>0.833333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ffdddff32e7ac6e2cbb5ae8471c1192c69ac6a94</th>\n",
       "      <td>10</td>\n",
       "      <td>12</td>\n",
       "      <td>0.833333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fff6d0f2fcaa034b7772dd1288c9222a6dcbbba0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fffe2ee6cd3b67aabb36f656f47f09591bc355e2</th>\n",
       "      <td>11</td>\n",
       "      <td>14</td>\n",
       "      <td>0.785714</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3153 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          sum  count     ratio\n",
       "revision_hash                                                 \n",
       "000e726af8464d94e7d02d307dde0e07e176d58c    2     18  0.111111\n",
       "001e2caa07a1dbfd92fbf86a26f30aaef23be5f5   51     66  0.772727\n",
       "002d7bf508ee59f23a04a9f21d02e9673dd38a71    4      4  1.000000\n",
       "002f8ec7aa01956b282ffd78062ee6fed6ab1eb9    1      1  1.000000\n",
       "004124ac5dbf5edbf925078652526267468821e7  165    166  0.993976\n",
       "...                                       ...    ...       ...\n",
       "ffc38b1f3b35ddfadb942240b9d2f933ec83d1be   62     63  0.984127\n",
       "ffc6adf7d493def070e26592bf828accbeecef2d    5      6  0.833333\n",
       "ffdddff32e7ac6e2cbb5ae8471c1192c69ac6a94   10     12  0.833333\n",
       "fff6d0f2fcaa034b7772dd1288c9222a6dcbbba0    1      1  1.000000\n",
       "fffe2ee6cd3b67aabb36f656f47f09591bc355e2   11     14  0.785714\n",
       "\n",
       "[3153 rows x 3 columns]"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "by_commit_eq_df = smartshark_vs_from_repos_df_merge_match_sel.groupby('revision_hash')['eq'].agg(['sum', 'count'])\n",
    "by_commit_eq_df['ratio'] = by_commit_eq_df['sum']/by_commit_eq_df['count']\n",
    "by_commit_eq_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "id": "0044dcdc-7b49-4cd1-8a5f-13b6bf28c5b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot(by_commit_eq_df['ratio'], rug=False, bins=20, kde=False, color='grey')\n",
    "\n",
    "ax = plt.gca()\n",
    "ax.set_title('Histogram of matched line annotations per commit')\n",
    "ax.set_xlabel('Percentage of matching line annotations')\n",
    "ax.set_ylabel('#Commits')\n",
    "#ax.set_xlim(0,1)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "id": "f9aaa34e-7a13-4e80-9679-d93c6b4ed246",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot(by_commit_eq_df['ratio'], rug=False, bins=20, kde=True)\n",
    "\n",
    "ax = plt.gca()\n",
    "ax.set_title('Histogram of matched line annotations per commit')\n",
    "ax.set_xlabel('Percentage of matching line annotations')\n",
    "ax.set_ylabel('#Commits')\n",
    "#ax.set_xlim(0,1)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "id": "e28a64ec-bdc1-48a1-9247-3cab6b9a665c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot(by_commit_eq_df['ratio'], rug=True, kind=\"ecdf\")\n",
    "\n",
    "ax = plt.gca()\n",
    "ax.set_title('Empirical cumulative distribution functions\\nof matched line annotations per commit')\n",
    "ax.set_xlabel('Percentage of matching line annotations')\n",
    "ax.set_ylabel('#Commits')\n",
    "#ax.set_xlim(0,1)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "id": "433f702e-ff7f-4615-8105-bc86f90d43e6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='ratio'>"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(x=by_commit_eq_df['ratio'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "id": "9192aed8-e7cc-425f-a3c6-605acf4bd9d5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='ratio'>"
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxenplot(x=by_commit_eq_df['ratio'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "id": "a3f2f573-17d1-4f78-aa3b-b19058340f3b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count      3153.000000\n",
       "mean          0.925410\n",
       "std           0.167948\n",
       "min           0.000000\n",
       "25%           0.929825\n",
       "50%           1.000000\n",
       "75%           1.000000\n",
       "90%           1.000000\n",
       "95%           1.000000\n",
       "99%           1.000000\n",
       "99.9%         1.000000\n",
       "99.99%        1.000000\n",
       "99.999%       1.000000\n",
       "max           1.000000\n",
       "Name: ratio, dtype: float64"
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "by_commit_eq_df['ratio'].describe(percentiles=[0.25,0.5,0.75,0.9,0.95,0.99,0.999,0.9999,0.99999])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "id": "29b503a6-3abe-49ea-83e1-3822d8687020",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    3153.000000\n",
       "mean        0.925410\n",
       "std         0.167948\n",
       "min         0.000000\n",
       "1%          0.000000\n",
       "5%          0.600000\n",
       "10%         0.797388\n",
       "20%         0.901408\n",
       "25%         0.929825\n",
       "30%         0.950000\n",
       "40%         0.979338\n",
       "50%         1.000000\n",
       "75%         1.000000\n",
       "max         1.000000\n",
       "Name: ratio, dtype: float64"
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "by_commit_eq_df['ratio'].describe(percentiles=[0.01,0.05,0.1,0.2,0.25,0.3,0.4,0.5,0.75])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "id": "5d81d781-ca46-4f8d-b03f-c12d2e5451b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.stats import bootstrap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "id": "af0d9393-bc6d-4a8a-8469-2eec1e229856",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.11111111, 0.77272727, 1.        , ..., 0.83333333, 1.        ,\n",
       "       0.78571429])"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "by_commit_eq_ratio_data = by_commit_eq_df['ratio'].values\n",
    "by_commit_eq_ratio_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "id": "1d662acf-96cd-4e49-98b5-8cd4350b1ba9",
   "metadata": {},
   "outputs": [],
   "source": [
    "rng = np.random.default_rng()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "id": "b226b9ca-9e62-4399-a3c5-6854bfdd4175",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "res.standard_error=np.float64(0.006101590221833957)\n",
      "ConfidenceInterval(low=np.float64(0.15814645912895386), high=np.float64(0.17819033287300207))\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "res = bootstrap((by_commit_eq_ratio_data,), np.std, confidence_level=0.9,\n",
    "                random_state=rng)\n",
    "fig, ax = plt.subplots()\n",
    "ax.hist(res.bootstrap_distribution, bins=25)\n",
    "ax.set_title('Bootstrap Distribution')\n",
    "ax.set_xlabel('statistic value (std)')\n",
    "ax.set_ylabel('frequency')\n",
    "\n",
    "print(f\"{res.standard_error=}\")\n",
    "print(res.confidence_interval)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "id": "c305cff3-836b-4ee3-a7e6-52a1437f93bf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "res.standard_error=np.float64(0.0030389531989620327)\n",
      "ConfidenceInterval(low=np.float64(0.9202660119001642), high=np.float64(0.9302888934296732))\n",
      "by_commit_eq_df['ratio'].mean()=np.float64(0.9254102422058237)\n",
      "by_commit_eq_df['ratio'].std()=np.float64(0.16794768506468252)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "res = bootstrap((by_commit_eq_ratio_data,), np.mean, confidence_level=0.9,\n",
    "                random_state=rng)\n",
    "fig, ax = plt.subplots()\n",
    "ax.hist(res.bootstrap_distribution, bins=25)\n",
    "ax.set_title('Bootstrap Distribution')\n",
    "ax.set_xlabel('statistic value (mean)')\n",
    "ax.set_ylabel('frequency')\n",
    "ax.grid()\n",
    "\n",
    "print(f\"{res.standard_error=}\")\n",
    "print(res.confidence_interval)\n",
    "print(f\"{by_commit_eq_df['ratio'].mean()=}\")\n",
    "print(f\"{by_commit_eq_df['ratio'].std()=}\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "id": "095e9abc-4d27-45e9-8896-758807d2cc66",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "res.standard_error=np.float64(0.0)\n",
      "ConfidenceInterval(low=np.float64(nan), high=np.float64(nan))\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jnareb/python-diff-annotator/.venv/lib/python3.12/site-packages/scipy/stats/_resampling.py:149: RuntimeWarning: invalid value encountered in scalar divide\n",
      "  a_hat = 1/6 * sum(nums) / sum(dens)**(3/2)\n",
      "/tmp/ipykernel_3965774/2321583984.py:1: DegenerateDataWarning: The BCa confidence interval cannot be calculated. This problem is known to occur when the distribution is degenerate or the statistic is np.min.\n",
      "  res = bootstrap((by_commit_eq_ratio_data,), np.median,\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "res = bootstrap((by_commit_eq_ratio_data,), np.median,\n",
    "                confidence_level=0.9, n_resamples=99999,\n",
    "                random_state=rng)\n",
    "fig, ax = plt.subplots()\n",
    "ax.hist(res.bootstrap_distribution, bins=25)\n",
    "ax.set_title('Bootstrap Distribution')\n",
    "ax.set_xlabel('statistic value (median)')\n",
    "ax.set_ylabel('frequency')\n",
    "\n",
    "print(f\"{res.standard_error=}\")\n",
    "print(res.confidence_interval)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0fd7f615-d813-409f-8914-e36aebda6165",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}