lectures/21_communication.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "![](img/330-banner.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "# Lecture 21: Communication\n",
    "\n",
    "UBC 2022 Summer\n",
    "\n",
    "Instructor: Mehrdad Oveisi"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.compose import ColumnTransformer, TransformedTargetRegressor\n",
    "from sklearn.dummy import DummyClassifier, DummyRegressor\n",
    "from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.linear_model import Ridge\n",
    "from sklearn.metrics import log_loss\n",
    "from sklearn.model_selection import (\n",
    "    GridSearchCV,\n",
    "    cross_val_score,\n",
    "    cross_validate,\n",
    "    train_test_split,\n",
    ")\n",
    "from sklearn.pipeline import Pipeline, make_pipeline\n",
    "from sklearn.preprocessing import (\n",
    "    MinMaxScaler,\n",
    "    OneHotEncoder,\n",
    "    OrdinalEncoder,\n",
    "    StandardScaler,\n",
    ")\n",
    "\n",
    "plt.rcParams[\"font.size\"] = 16"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "## Learning objectives\n",
    "\n",
    "- When communicating about applied ML, tailor an explanation to the intended audience.\n",
    "- Apply best practices of technical communication, such as bottom-up explanations and reader-centric writing.\n",
    "- Given an ML problem, analyze the decision being made and the objectives.\n",
    "- Avoid the pitfall of thinking about ML as coding in isolation; build the habit of relating your work to the surrounding context and stakeholders.\n",
    "- Interpret a confidence score or credence, e.g. what does it mean to be 5% confident that a statement is true.\n",
    "- Maintain a healthy skepticism of `predict_proba` scores and their possible interpretation as credences.\n",
    "- Be careful and precise when communicating confidence to stakeholders in an ML project.\n",
    "- Identify misleading visualizations."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Lecture plan \n",
    "\n",
    "- Announcements\n",
    "- Motivation (~5 min)\n",
    "- Activity: explaining GridSearchCV (~15 min)\n",
    "- Principles of good explanation (~15 mins)\n",
    "- ML and decision-making (~5 mins)\n",
    "- Break (5 min)\n",
    "- Confidence and predict_proba (~20 mins) \n",
    "- Credence activity (~10 mins) (time permitting)\n",
    "- Misleading visualization (~10 min)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Announcements\n",
    "\n",
    "- hw6 grades posted\n",
    "- hw8 (last one) deadline on December 7th at 11:59pm\n",
    "- I will post information about the format of the final sometime this weekend. \n",
    "- Almost there, hang in there!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "## Attribution\n",
    "\n",
    "- The first part of this lecture is adapted from [DSCI 542](https://github.com/UBC-MDS/DSCI_542_comm-arg), created by [David Laing](https://davidklaing.com/).\n",
    "- The visualization component of this lecture benefitted from discussions with [Firas Moosvi](http://firas.moosvi.com/) about his course, [DSCI 531](https://github.com/UBC-MDS/DSCI_531_viz-1)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Why should we care about effective communication?\n",
    "\n",
    "- Most ML practitioners work in an organization with >1 people.\n",
    "- There will very likely be stakeholders other than yourself.\n",
    "- Those people need to understand what you're doing because:\n",
    "  - their state of mind may change the way you do things (see below)\n",
    "  - your state of mind may change the way they do things (interpreting your results)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "ML suffers from some particular communication issues:\n",
    "- overstating one's results / unable to articulate the limitations\n",
    "- unable to explain the predictions\n",
    "- Can we trust test error?\n",
    "- Why did CatBoost make that prediction?\n",
    "- What does it mean if `predict_proba` outputs 0.9?\n",
    "\n",
    "These issues are there because these things are actually very hard to explain!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Activity: explaining `GridSearchCV` (15 min)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "Below are two possible explanations of `GridSearchCV` pitched to different audiences. Read them both and then follow the instructions at the end."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Explanation 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Machine learning algorithms, like an airplane's cockpit, typically involve a bunch of knobs and switches that need to be set.\n",
    "\n",
    "![](https://i.pinimg.com/236x/ea/43/f3/ea43f3c7f3a8c92d884ce012c77628fd--cockpit-gauges.jpg)\n",
    "\n",
    "For example, check out the documentation of the popular random forest algorithm [here](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassifier.html). Here's a list of the function arguments, along with their default values (from the documentation):\n",
    "\n",
    "> class sklearn.ensemble.RandomForestClassifier(n_estimators=100, criterion='gini', max_depth=None, min_samples_split=2, min_samples_leaf=1, min_weight_fraction_leaf=0.0, max_features='auto', max_leaf_nodes=None, min_impurity_decrease=0.0, min_impurity_split=None, bootstrap=True, oob_score=False, n_jobs=None, random_state=None, verbose=0, warm_start=False, class_weight=None, ccp_alpha=0.0, max_samples=None)\n",
    "\n",
    "Holy cow, that's a lot of knobs and switches! As a machine learning practitioner, how am I supposed to choose `n_estimators`? Should I leave it at the default of 100? Or try 1000? What about `criterion` or `class_weight` for that matter? Should I trust the defaults?\n",
    "\n",
    "Enter [GridSearchCV](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.html) to save the day. The general strategy here is the choose the settings that perform best on the specific task of interest. So I can't say `n_estimators=100` is better than `n_estimators=1000` without knowing what problem I'm working on. For a specific problem, you usually have a numerical score that measures performance. `GridSearchCV` is part of the popular [scikit-learn](https://scikit-learn.org/) Python machine learning library. It works by searching over various settings and tells you which one worked best on your problem. \n",
    "\n",
    "The \"grid\" in \"grid search\" comes from the fact that tries all possible combinations on a grid. For example, if you want it to consider setting `n_estimators` to 100, 150 or 200, and you want it to consider setting `criterion` to `'gini'` or `'entropy'`, then it will search over all 6 possible combinations in a grid of 3 possible values by 2 possible values: \n",
    "\n",
    "|                    | `criterion='gini'` | `criterion='entropy'` |\n",
    "|----------------------|--------|---------|\n",
    "| `n_estimators=100` |    1     |    2     |\n",
    "| `n_estimators=150` |    3     |    4     |\n",
    "| `n_estimators=200` |    5     |    6     |\n",
    "\n",
    "Here is a code sample that uses `GridSearchCV` to select from the 6 options we just mentioned. The problem being solved is classifying images of handwritten digits into the 10 digit categories (0-9). I chose this because the dataset is conveniently built in to scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'criterion': 'gini', 'n_estimators': 100}"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# imports\n",
    "from sklearn import datasets\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "# load a dataset\n",
    "data = datasets.load_digits()\n",
    "X = data[\"data\"]\n",
    "y = data[\"target\"]\n",
    "\n",
    "# set up the grid search\n",
    "grid_search = GridSearchCV(\n",
    "    RandomForestClassifier(random_state=123),\n",
    "    param_grid={\"n_estimators\": [100, 150, 200], \"criterion\": [\"gini\", \"entropy\"]},\n",
    ")\n",
    "\n",
    "# run the grid search\n",
    "grid_search.fit(X, y)\n",
    "grid_search.best_params_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see from the output above, the grid search selected `criterion='gini', n_estimators=100`, which was one of our 6 options above (specifically Option 1).\n",
    "\n",
    "By the way, these \"knobs\" we've been setting are called [_hyperparameters_](https://en.wikipedia.org/wiki/Hyperparameter_(machine_learning) and the process of setting these hyperparameters automatically is called [_hyperparameter optimization_](https://en.wikipedia.org/wiki/Hyperparameter_optimization) or _hyperparameter tuning_."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "~400 words, not including code."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "<br><br><br><br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Explanation 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "https://medium.com/datadriveninvestor/an-introduction-to-grid-search-ff57adcc0998\n",
    "\n",
    "~400 words, not including code."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "<br><br><br><br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Discussion questions:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- What do you like about each explanation?\n",
    "- What do you dislike about each explanation?\n",
    "- What do you think is the intended audience for each explanation?\n",
    "- Which explanation do you think is more effective overall for someone on Day 1 of CPSC 330?\n",
    "- Each explanation has an image. Which one is more effective? What are the pros/cons?\n",
    "- Each explanation has some sample code. Which one is more effective? What are the pros/cons?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After you're done reading, take ~5 min to consider the discussion questions above. Paste your answer to **at least one** of the above questions in the [Google doc](https://docs.google.com/document/d/1y4YMSAWQGRFFn8ig0pAs5TJR5h2jKnvtY97HN-rEkyE/edit?usp=sharing) under the appropriate question heading."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Principles of good explanations (15 min)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Concepts *then* labels, not the other way around\n",
    "\n",
    "The first explanation start with an analogy for the concept (and the label is left until the very end):\n",
    "\n",
    "> Machine learning algorithms, like an airplane's cockpit, typically involve a bunch of knobs and switches that need to be set.\n",
    "\n",
    "In the second explanation, the first sentence is wasted on anyone who doesn't already know what \"hyperparameter tuning\" means:\n",
    "\n",
    "> Grid search is the process of performing hyper parameter tuning in order to determine the optimal values for a given model. \n",
    "\n",
    "The effectiveness of these different statements depend on your audience. \n",
    "\n",
    "See [this video](https://twitter.com/ProfFeynman/status/899963856549625858?s=20): \"I learned very early the difference between knowing the name of something and knowing something.\" -Richard Feynman."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Bottom-up explanations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The [Curse of Knowledge](https://en.wikipedia.org/wiki/Curse_of_knowledge) leads to *top-down* explanations:\n",
    "\n",
    "![](img/top_down.png)\n",
    "\n",
    "- When you know something well, you think about things in the context of all your knowledge. \n",
    "- Those lacking the context, or frame of mind, cannot easily understand. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "There is another way: *bottom-up* explanations:\n",
    "\n",
    "![](img/bottom_up.png)\n",
    "\n",
    "When you're brand new to a concept, you benefit from analogies, concrete examples and familiar patterns.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### New ideas in small chunks\n",
    "\n",
    "The first explanation has a hidden conceptual skeleton:\n",
    "\n",
    "1. The concept of setting a bunch of values.\n",
    "2. Random forest example.\n",
    "3. The problem / pain point.\n",
    "4. The solution.\n",
    "5. How it works - high level.\n",
    "6. How it works - written example.\n",
    "7. How it works - code example.\n",
    "8. The name of what we were discussing all this time."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Reuse your running examples\n",
    "\n",
    "Effective explanations often use the same example throughout the text and code. This helps readers follow the line of reasoning."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Approach from all angles\n",
    "\n",
    "When we're trying to draw mental boundaries around a concept, it's helpful to see examples on all sides of those boundaries. If we were writing a longer explanation, it might have been better to show more, e.g.\n",
    "\n",
    "- Performance with and without hyperparameter tuning. \n",
    "- Other types of hyperparameter tuning (e.g. `RandomizedSearchCV`)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### When experimenting, show the results asap\n",
    "\n",
    "The first explanation shows the output of the code, whereas the second does not. This is easy to do and makes a big difference."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Interesting to you != useful to the reader (aka it's not about you)\n",
    "\n",
    "Here is something which was deleted from the explanation:\n",
    "\n",
    "> Some hyperparameters, like `n_estimators` are numeric. Numeric hyperparameters are like the knobs in the cockpit: you can tune them continuously. `n_estimators` is numeric. Categorical hyperparameters are like the switches in the cockpit: they can take on (two or more) distinct values. `criterion` is categorical. \n",
    "\n",
    "It's a very elegant analogy! But is it helpful?\n",
    "\n",
    "And furthermore, what is my hidden motivation for wanting to include it? Elegance, art, and the pursuit of higher beauty? Or _making myself look smart_? So maybe another name for this principle could be **It's not about you.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## ML and decision-making (5 min)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- There is often a wide gap between what people care about and what ML can do.\n",
    "- To understand what ML can do, let's think about what **decisions** will be made using ML. \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Decisions involve a few key pieces\n",
    "\n",
    "- The **decision variable**: the variable that is manipulated through the decision.\n",
    "  - E.g. how much should I sell my house for? (numeric)\n",
    "  - E.g. should I sell my house? (categorical)\n",
    "- The decision-maker's **objectives**: the variables that the decision-maker ultimately cares about, and wishes to manipulate indirectly through the decision variable.\n",
    "  - E.g. my total profit, time to sale, etc.\n",
    "- The **context**: the variables that mediate the relationship between the decision variable and the objectives.\n",
    "  - E.g. the housing market, cost of marketing it, my timeline, etc."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### How does this inform you as an ML practitioner?\n",
    "\n",
    "Questions you have to answer:\n",
    "\n",
    "- Who is the decision maker?\n",
    "- What are their objectives?\n",
    "- What are their alternatives?\n",
    "- What is their context?\n",
    "- What data do I need?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Break (5 min)\n",
    "\n",
    "![](img/eva-coffee.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Confidence and `predict_proba` (20 min)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- What does it mean to be \"confident\" in your results?\n",
    "- When you perform analysis, you are responsible for many judgment calls.\n",
    "- [Your results will be different than others'](https://fivethirtyeight.com/features/science-isnt-broken/#part1).\n",
    "- As you make these judgments and start to form conclusions, how can you recognize your own uncertainties about the data so that you can communicate confidently?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### What does this mean for us, when we're trying to make claims about our data?\n",
    "\n",
    "Let's imagine that the following claim is true:\n",
    "\n",
    "> Vancouver has the highest cost of living of all cities in Canada.\n",
    "\n",
    "Now let's consider a few beliefs we could hold:\n",
    "\n",
    "1. Vancouver has the highest cost of living of all cities in Canada. **I am 95% sure of this.** \n",
    "2. Vancouver has the highest cost of living of all cities in Canada. **I am 55% sure of this.** \n",
    "\n",
    "The part is bold is called a [credence](https://en.wikipedia.org/wiki/Credence_(statistics)). Which belief is better?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "But what if it's actually Toronto that has the highest cost of living in Canada?\n",
    "\n",
    "1. Vancouver has the highest cost of living of all cities in Canada. **I am 95% sure of this.** \n",
    "2. Vancouver has the highest cost of living of all cities in Canada. **I am 55% sure of this.** \n",
    "\n",
    "Which belief is better now?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "**Conclusion: We don't just want to be right. We want to be confident when we're right and hesitant when we're wrong.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### What do credences mean in practical terms?\n",
    "\n",
    "One of two things:\n",
    "\n",
    "* **I would accept a bet at these odds.** 99% sure means, \"For the chance of winning \\\\$1, I would bet \\\\$99 that I'm right about this.\" 75% sure means, \"For the chance of winning \\\\$25, I would bet \\\\$75 that I'm right about this.\"\n",
    "* **Long-run frequency of correctness.** 99% sure means, \"For every 100 predictions I make at this level of confidence, I would expect only 1 of them to be incorrect.\" 75% sure means, \"For every 100 predictions I make at this level of confidence, I would expect about 25 of them to be incorrect.\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### It's easy enough to evaluate how good we are at being right...\n",
    "\n",
    "But if we want to evaluate *how good we are at knowing how right we are*?\n",
    "\n",
    "We would need to keep of not just the correctness of our predictions, but also our confidence in those predictions.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### What does this have to do with applied ML?\n",
    "\n",
    "- What if you `predict` that a credit card transaction is fraudulent?\n",
    "  - We probably want `predict_proba` a lot of the time.\n",
    "- What if `predict_proba` is 0.95 in that case?\n",
    "  - How confident are YOU?\n",
    "- What if you forecast that avocado prices will go up next week? \n",
    "  - How confident are you there?\n",
    "- Or what if you predict a house price to be \\\\$800k? \n",
    "  - That is not even a true/false statement."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Loss functions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When you call `fit` for `LogisticRegression` it has these same preferences: correct and confident > correct and hesitant > incorrect and hesitant > incorrect and confident."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import log_loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- This is a \"loss\" or \"error\" function like mean squared error, so lower values are better.\n",
    "- When you call `fit` it tries to minimize this metric."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Correct and 95% confident:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.05129329438755058"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_loss(y_true=np.array([0]), y_pred=np.array([[0.95, 0.05]]), labels=(0, 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Correct and 55% confident:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5978370007556204"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_loss(y_true=np.array([0]), y_pred=np.array([[0.55, 0.45]]), labels=(0, 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Incorrect and 55% confident:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7985076962177716"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_loss(y_true=np.array([0]), y_pred=np.array([[0.45, 0.55]]), labels=(0, 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Incorrect and 95% confident:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.995732273553991"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_loss(y_true=np.array([0]), y_pred=np.array([[0.05, 0.95]]), labels=(0, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEQCAYAAABBQVgLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAAt0ElEQVR4nO3dd7xcVbn/8c9zenISwklI6CGJqBSl3UjXEBQEpSnYC+Uq4k8B5aqIKCI2uveCFdSLV8AG1wuKBTAJRWpA6d2QQAiQRtpJTn1+f6w1yZ5yzszOzJyZnPN9v17zmpldnz1zzn5m7bXXWubuiIiIDKah1gGIiEj9U7IQEZGilCxERKQoJQsRESlKyUJERIpqqnUA1bDFFlv4lClTah2GiMgm5YEHHlji7hMLzRuWyWLKlCnMnTu31mGIiGxSzGz+QPN0GUpERIpSshARkaKULEREpCglCxERKUrJQkREilKyEBGRopQsRESkKCWLhM7uXi69+Sn+sWB5rUMREakrShYJa7v7uGzWszyycEWtQxERqStKFgVoPCgRkWxKFglmVusQRETqkpJFARpqVkQkm5JFgsoVIiKFKVkUoHKFiEg2JYsEVVmIiBSmZFGAqixERLIpWSSYai1ERApSsihABQsRkWxKFkkqWIiIFKRkUYDaWYiIZFOySNDdUCIihSlZiIhIUUoWCSpYiIgUpmRRgKosRESyKVkkqNdZEZHCmtIsbGYtwF7ANsAoYAnwlLs/X/nQasfV0kJEJEvRZGFmjcB7gE8AM4AWsi/vu5ktBH4FXOnuz1Yj0KGgcoWISGGDXoYys2OBJ4GrgS7gq8AhwO7AG4B9gQ8D1wHHAE+Y2ZVmtmUVY6461VmIiGQrVrK4HLgQuMrdXxtgmfuA3wBnmNk+wJnAJ4FvVSrIoaIqCxGRwopVcE9z9/8cJFFkcfd73f29wEWF5pvZO81slpm9bGZdZvaimf3WzHbJWa7DzH5qZkvMbI2Z3Wpmby7piCpABQsRkWyDJgt3XwehYtvMTjezN5WyUXfvGmDWeOAB4LPAocBZwK7APWa2Q9yXATcChwGnAscCzcBsM9uulP1vLPU6KyJSWEl3Q7l7t5mdD7yznJ25+68IFeHrmdl9hHqR44BLgKOAA4GD3X12XOZuYB7wJeC0cmIoLc5q70FEZNOSpp3FE8C0KsSwND73xOejgJcyiQLA3VcAfwCOrsL+11OdhYhIYWmSxTnA1ypRd2BmjfHS1uuBnwAvA7+Os3cFHi2w2mPAZDMbU+7+i1E7CxGRbGka5Z0JjAH+YWbPA4vIrgt2d59R4rbuBf4tvn6WcMnp1fh+PPB8gXWWxecOYHXuTDM7GTgZYPLkySWGISIipUhTsugDHgfuAF4AeuO0zKM/xbY+xoY2GiuBW8xsSpxnFL4hadCLRO5+hbtPd/fpEydOTBGKiIgUU3LJwt0PqtRO3f2J+PJeM/szoSTxZeAUQglifIHVOuLz8krFMXB81d6DiMimpeYdCcY2HM8CO8ZJjxHqLXLtAixw97xLUJWiCm4RkcJSJQsz29bMLjWzuWY2L9Puwsw+F1tvpxa7BtkJeC5OuhHY1sxmJJbZDDgyzhMRkSFW8mUoM9uVUF/RB9wN7EnoVBBgB2BvQh3EYNv4PfAg8DChruINwOcJ9R+XxMVujNu/2sy+SLjsdBahzuLCUuPdGGqUJyJSWJqSxSWEthZTgfeSXeF8F6HCuph7CB0O/gK4CTgDuA3Yw92fBnD3fuAI4Bbgh8DvCQlqpru/kCLejeaqtBARyZLm1tkDgQ+5++rYbXnSK8BWxTbg7hcAF5Sw3DLgpPgYMqqzEBEpLE3JYrBbY7cA1pYZS91QwUJEJFuaZHEfcOIA894P/L38cGpLBQsRkcLSXIb6JnCrmd0MXEtoOPcOMzudMJLe26oQX02oYCEikq3kkoW730aonJ4K/JzwQ/x84K3AMe5+bzUCHEqmSgsRkYLSlCxw95uAm8xsR2ASsNTdn6pKZDWkOgsRkWwllyzM7Bwz2wbA3Z9197syicLMtjazc6oV5FBRuUJEpLA0FdxfBwYaqW6bOH9YUBflIiLZ0iSLwX54dwADDaW6yVCVhYhIYYPWWZjZQcDBiUmfMrMjchYbBbyb0AHgsKA6CxGRbMUquGcAX42vncLtLLoJ41ycXsG4akJ3Q4mIFDboZSh3/4a7N7h7A+Ey1L6Z94lHm7vv5e53DU3I1aeChYhItjR1FjMJJYg8ZtZuZsOmUZ6IiGRLkyxmEQYgKmQnYHb54dQJVVqIiGSp1N1QrYRuxDd5qrYQEclX7G6oKcC0xKTpZjYmZ7FRhK7EF1Q2tNpRuUJEJFuxu6GOJzS28/i4nOwShsf3vcBnqhHgUFPBQkQkX7FkcRUwh3AOnUVICLmV3F3A03HAomFBVRYiItkGTRbuPh+YD2BmM4EH3X3VUARWK2prISKSr+ReZ2MX5ZjZboSxKyYAP3H3l2MvtK8Ml0SivqFERLKVnCzMrBW4Gngv4bKUA38AXgYuBJ4GvlyFGIeUyhUiIvnS3Dr7beAdwMeALck+r/4ZeGcF46op1VmIiGRLM/jRh4Cvuvu1ZtaYM28eMKViUdWQqixERPKlKVlMAJ4YZDut5YdTH1SwEBHJliZZzAP2G2De3sCwGF7VVGshIpInTbL4H+DLZvYRoCVO83hL7eeBn1c6uFpRnYWISLY0yeJC4Cbgl0CmAd6dwK3AX9z98grHVhsqWIiI5EnTzqIP+KCZ/QA4DJgILCUkituqFF9NqJ2FiEi2NHdDAeDudwB3VCGWuqCChYhIvtTJItZR7AdsCywE7nL3ORWOq7ZUsBARyZKmBfd44HeEEfP6geVAR5hlc4D3DYfOBNXOQkQkX5oK7suAtwAfBUa5+0TCWBYfB6YD/1X58GpDBQsRkWxpLkMdCZzl7tdmJrh7D3BNLHV8q9LB1YLaWYiI5EtTsugDnhlg3lMMk2FVAVwNLUREsqRJFjcAHxhg3geB/ys7mjqgOgsRkXzFxuA+OPH2D8B/mtlNhIruVwi9z74f2BU4vVpBDjUVLEREshWrs7iVDeNsZ563Aw4vsOz1QG5vtJscFSxERPIVSxYzhySKOqOChYhItmJjcA+rbjxKoTG4RUTypangHjFUZyEikk3JIofKFSIi+ZQsClCvsyIi2ZQscqloISKSR8miANVZiIhkKytZmNkuZnasmW1TqYBqTQULEZF8JScLM/u+mf048f69wEOE1tyPm9lbStjGcWZ2vZnNN7O1ZvaUmX3XzMbmLNdhZj81syVmtsbMbjWzN6c4LhERqaA0JYvDgbsS778B/BHYHbgP+HoJ2/gCocPBrxCGZv0R8GngFjNrgDA4BnBjnH8qcCzQDMw2s+1SxLtR1M5CRCRfmi7KtwKeB4gn7V2Bf3f3R8zsMuBnJWzjSHdfnHh/m5ktA34BHATMAo4CDgQOdvfZcX93A/OALwGnpYh5o6jXWRGRbGlKFmuBMfH1DGAlMDe+Xw2MLbRSUk6iyLg/Pm8bn48CXsokirjeCkJHhkeniHejqGAhIpIvTbJ4EPiMmb0J+Axwi7v3x3lTgUUbGcOM+PxEfN4VeLTAco8Bk81sTIF5FaVyhYhItjTJ4mxgX0Kl9huBbybmHUOot0jFzLYFzgNudfdMKWU8YXzvXJnxvTsG2NbJZjbXzOYuXlyoACMiIhur5DoLd7/fzCYDOwHPuPvKxOwrGHgUvYJiCeEGoBc4MTmLwj/uB71A5O5XxDiYPn36RhcOdBVKRCRfmgpu3H0N8ECB6Tel2Y6ZtRHueJoGzHD3FxOzlxFKF7kyJYpCpY6KUv22iEi2YiPlfRy4yd2XxteDcvf/KbaMmTUTBkraG3iHuz+Ss8hjwKEFVt0FWODuq4vtoxy6dVZEJF+xksVVhHqKpfH1YBwYNFnEthTXAG8H3u3u9xRY7EbgRDObkRlPw8w2A44Eri0SQ0WoI0ERkWzFkkXyLqepFdjfD4D3Ad8G1pjZvol5L8bLUTcCdwNXm9kXCZedziJUJ1xYgRgGpXKFiEi+YiPlzS/0ugyZsbvPjo+kbwDnunu/mR0BXAz8EGgjJI+Z7v5CBWIoSnUWIiLZUlVwl8vdp5S43DLgpPgYUqqyEBHJpy7KC1DBQkQkm5JFHhUtRERyKVkUoDoLEZFsJSULM2s0s93NbGK1A6o11VmIiOQrtWThhB5m96xiLHVERQsRkaSSkkXsXfYFoL264dSeChYiIvnS1Fn8BPicmbVUK5h6oToLEZFsadpZjAVeB/zLzP5CaNmdPK26u5cytGpdU52FiEi+NMniK4nXhRrLOaWNw133VLIQEcmWZjyLEXGbranWQkQkz4hIAGmp11kRkWypkoUFR5nZxWb232a2Q5w+w8y2qU6IQ0t1FiIi+Uq+DGVmHcCfgH2AlYQK78uB+cAnCSPcnVaFGIec6ixERLKlKVlcBGwPHABsQXaThFsJAxpt8lSwEBHJl+ZuqKOBL7j73WbWmDNvASGRDAsqWIiIZEtTshgDLBxgXhvD5Ee5xuAWEcmXJlk8BRw6wLwZwCPlh1MfVGchIpItzWWoHwA/MLMVwLVx2uZmdiLwWeDkSgcnIiL1IU2jvCvN7HWEsbLPi5NvAfqBC939mirEVxNqZyEiki3VGNzu/mUz+xFwCDAJWArc4u7/qkZwtaAqCxGRfKmSBYC7zwd+WoVY6ocKFiIiWdI0yrsHmAXMAe5w97XVCqqWVLIQEcmX5m6o54Djgb8Ay83sDjM7z8wOMrPW6oRXGypYiIhkKzlZuPtH3H1bYBfgDMJ4FqcAfyMkj79VJ8ShpV5nRUType511t2fdPcfEsa0OIlwaaoNOKiyodWOq6GFiEiWNHUWbcCBwEzgYODfgDXAncAXCEljk6c6CxGRfGnuhlpOuJx/B3AD8Dngfnfvr0JcNaVyhYhItjSXoVYDrcCWhDYWk4D2agRVSypYiIjkS9OCe6KZ7caGy1AnAmPM7EFgNjDL3W+uTphDS1UWIiLZUlVwu/vD7v5f7n40MIEwhsVK4EvAn6sQ35BTr7MiIvlSteA2s2ZgX0LJYiZh1LxW4FVCY71hQQULEZFsae6GugXYDxhNGEL1NuCLwGx3f6w64Q09lStERPKlKVmsA75GqJ94yIdxY4RhfGgiIhslTQX3kdUMpG6oaCEikid1r7NmdgRhZLzxhC7K57j7nyodWC2pXCEiki1NncVY4I/AW4FeQqKYAPyHmd0BHOHuq6sS5RBSwUJEJF+aW2e/A+wFfAwY5e5bA6OAj8fp36l8eDWiooWISJY0yeJY4Kvufo279wG4e18cTvVrcf4mT+0sRETypUkWE4DHB5j3eJw/LGgMbhGRbGmSxTzgiAHmvSvO3+SpXCEiki/N3VA/AS4xszHANYTBj7YCPgh8gjAg0rCgZhYiItnStLP4nplNBD4PnBAnG9AFnO/u/1X58IaeqixERPKlamfh7l8xs4sI/UONJ3T7cY+7L69GcLWikoWISLbUjfJiYhgWPcwWojG4RUTyDZoszOxtaTbm7reXF0590N1QIiLZipUs5lBaEzWLyzUWXdBsO+BMYDqwO6Fh31R3fz5nuQ7gIuCYuMzdwOfd/ZES4hERkQoqlixmVmGfOwLvBx4gjOd9aO4CFlrG3QhMBU4ljP99FjDbzPZw9xerEFfct+osRERyDZos3P22KuzzdnffEsDMPkGBZAEcBRwIHOzus+OydxPacnwJOK0KcYmIyABSDataCe7eX8JiRwEvZRJFXG8F8Afg6GrFJiIihQ2aLMzsBjPbs9SNmVmbmZ1hZqeUGdeuwKMFpj8GTI4NA6tGV6FERLIVK1ksAO4xs3vN7DQz28vMsi5dmdk2ZnaMmf2M0Kr7JODBMuMaT6inyLUsPnfkzjCzk81srpnNXbx48UbvWB0JiojkGzRZuPupwC7AfcC5wP3AOjNbZmaLzGwd8ALwv4TSwOeA3dz9vjLjytxdVWj6QLFe4e7T3X36xIkTy9q5KrhFRLIVbZTn7s8Bp5rZfwD7AfsA2wBthAGQniRUWs+vYFzLCKWLXJkSRdVajKtcISKSL03fUN3AbfFRbY9R+C6pXYAF1R+RT0ULEZGkIb8bqkQ3Atua2YzMBDPbDDgyzqsaVVmIiORL3TdUJZjZcfHlv8Xnw81sMbA4tu24kdBi+2oz+yIbGuUZcGG141OdhYhItpokC+B3Oe9/GJ9vAw5y934zOwK4OM5rIySPme7+QjUDU8lCRCRfTZKFuxc9Jbv7MsJtuCdVP6KcfQ/1DkVE6ly91lnUjLooFxHJV3ayMLMJlQiknrgqLUREspScLMzsk7GyOfP+zWb2IvBqbDm9VVUiHGKqsxARyZemZHEqsDbx/lLgNUKr7XHAeRWLqsZUrhARyZamgnsyobU2ZjYOmAEc4+5/MrOlwHerEN+QU8FCRCRfmpJFI5DpXvxAwg/wOfH9C8CkyoVVW6qyEBHJliZZPAO8O77+IHCXu3fG99uwoUfYTZsqLURE8qS5DHUx8EszO57Qod/7EvNmAg9XMrBaUsFCRCRbmo4ErzWzBYReZ+9399sTs1+hyn02DRWVK0RE8qVqwe3udwJ3Fpj+9YpFVGONDUZvXykjv4qIjBxp2lnsH/tryryfYGa/MrNHzOxiM2usTohDq2N0M6919tQ6DBGRupKmgvt8NvQSC3AR8C7gaeDTwFcqGFfNbD66heWd3bUOQ0SkrqRJFjsDcwHMrBk4Dvi8ux8LnA18uPLhDb3x7S0sW6NkISKSlCZZjAFWxtd7A+3AH+P7BwmN9jZ5E9pb6OrtZ8VaXYoSEclIkywWArvH14cDj7r7q/F9B9BZcK1NzNQt2gGYt2RNjSMREakfaZLFr4DvmNl1wBnA1Yl5exEa7W3ypk0cA8C/Fld5mG8RkU1ImltnzwXWAfsSKru/l5i3O/mj322SJo8fTWOD8ZyShYjIemka5fUB3x5g3jGVCqjWWpoamLZFO4+9tLL4wiIiI0TqYVXN7E2EHmfHA0uB29390UoHVkv7TpvA9Q++SHdvPy1NGkxQRCRNo7wmM7saeAi4HPgG8H3gITP75XBplAdwwI4T6OzuY+7zw6NvRBGRcqX52fx14P3AOcBUYFR8Pgf4QHweFma8YRJj25r47dwXah2KiEhdSJMsPgp8092/7e7z3b0rPn8b+Bbw8eqEOPRGtTRyzB7b8qdHX2bp6q5ahyMiUnNpksU2wN0DzLsrzh82jt9/B3r7+rl81rO1DkVEpObSJIuXgAMGmLd/nD9s7DhpLB94y2Suvmc+D7/4Wq3DERGpqTTJ4hrgbDP7mplNM7NRZjbVzM4i9A31y+qEWDtnHvZGJo1t5bPX/kPdf4jIiJYmWZwLXEe4C+oZYDXwLKHtxXXAeZUOrtY2H93CZR/ak0Ur1nLSVfezpqu31iGJiNREycnC3Xvd/cPAm4HPEu5++izwJnf/iLsPyzPp9CnjueyDe/KPBcv58JX38MrKdbUOSURkyKVucebuj7n7j+JdUT9y98fN7B1mNmzG4M51+Ju35icfm84zr67myMvvZM5TrxZfSURkGKlU8+RxwK4V2lZdOmSXLbn+0/szblQzJ/z3/Xzhdw+plCEiI4b6skhh56034w+nHsinD3odN/xzITMums1Ff32SxavUFkNEhjcli5Tamhs587Cd+NsZB3HILlvxwznPccD5s/ji7x7i0YUrcPdahygiUnFWiZObmR0L/Nbd66J/qOnTp/vcuXOHZF//Wrya//7781z3wIus7enj9ZPG8J69tuXI3bZh+/GjhyQGEZFKMLMH3H16wXmDJQszm1biPg4HLhuJySJjRWcPf3j4Jf7vHwuZO385ADttNZaDd5rE23fekt23G0dTowpyIlK/ykkW/UApRQ8DfCQni6QFSzu5+fGXufWJV7j/+eX09TtjWpuYPqWDfadNYJ+p43nTtuNoVvIQkTpSTrI4Ps2O3P0XKWOriloni6QVnT3c/sxi7vnXUu6dt4xnXw0j8LU1N7DrNuN487bj2G278Ji6xRgaG6zGEYvISLXRyWJTVU/JIteS1V3cN28Zc59fzqMLV/DoSyvo7O4DoL2lkTdsNZbXTxrDG7Ycy+u3DK+3HteGmZKIiFSXkkUd6+t3/rV4NQ+/uIJHFq7gqZdX8cyrq1iyunv9MmNbm3jdpDFMmTCayRPamTJhNDtMGM3k8e1sMaZFiUREKkLJYhO0bE03z7yyiqdfXc2zr6zimVdXM39pJ4tWrKU/8ZW1tzSy/fiQPLYeN4ptNm9b/7zN5qOYNLZNl7ZEpCSDJYvUY3DL0Bjf3sI+0yawz7QJWdO7evt4cflaFiztZP7SNcxf1sn8pZ08t3gNdz6zhDXxklZGY4Ox5dhWtt58FFuPa2PrcW1MHNsaHmM2vN58VDMNSioiMgAli01Ma1Mjr5s4htdNHJM3z91Zua6XRSvWsui1dbyU8/zowhXc+sQrrOvpz1u3qcHYYkwrW4xtYeKY1vVJZHx7K+Pbm+kY3cL49hY6RrfQ0d5Ce0ujLn+JjCBKFsOImTFuVDPjRjWz01abFVzG3Vnd1cviVV0sXtXFktXdLF61jsWru9ZPW7y6i8cXrWTJ6m76+gtfpmxpbKAjmUTaWxg/uoWO0c10tLesj2OzzHNbeG5rblCSEdkEKVmMMGbG2LZmxrY1M61A6SSpv99Zua6HZWu6Wd7Zw/I13Szr7M56zkx/YtFKlq/p5rW1PQxWDdbcaOsTx9j1iaQpL6lsNqqJsW3NjGltZExrM+2tjYyNz2rcKDL0lCxkQA0NxuajW9h8dEvJ6/T1OyvW9rBibQ8r1/awcl3mdW94Xhemh9dh2ovLOtcv19NX/IaLtuYGxrTGRNLWRHtLE2PbmmhvbWJMaxNj2poY0xKe21ubGNsa57U1MbqlkdHNTYxqaaS9tZG2pkbV1YiUQMlCKqqxwRjfHi5NpeXurOvpX59sVnf1srqrlzVdvaxe17v+/frHujBvVVcvi1asW7/sqnW9dPXm18sMZFRzI6NbGhnVEp5Ht8Sk0tLIqJYmRjc3rk8uo1uacpZvoj3xenRLI63NDYxqbqStuVGt9GXYqNtkYWbbA98DDiF0J3Ir8Dl3X1DTwKRqzIxR8cS71bi2srbV09e/PnGs6Q6JZVVXL2u7++js7mNtdy+d8XVnfJ2Z19nTR2dXL6919rC2J87vCtMHqsMZSGOD0dbUwKiWRlqbGmlrbqCtuXF9MmlrbqB1/fsG2prC9LB8Q1wmMT/rOT4S29dt0lItdZkszGw0MAvoAo4n9E/1LWC2me3m7mtqGZ/Uv+bGhtSX0Ipxd7r7+jcklUTCyUxb091LV08f63r6WdfTx7rePtZ297Out491PX109fSztie87uzuZdmauFxPH+t6w+u1PX2D1vsMprHBaG1qoLWpgZamBlqbGhOvw/v1r5sbaWlsoLU5f/nWgss30NLYWHD55PabG003MQxDdZksgE8C04A3uvuzAHHY1meATwGX1jA2GaHMLJ4cG9m8ir3PZ5LSup5+umLyWJ98Eu+7YgJa2x0STXdvmNbV0093Xz9dcZkNr8Myr3V2r3/dtf7Rt35aJWQSSEtjeG5ubKC50WhpaqSl0WjOmh4STXPO9Mz6zY0NNDdZzrYy8wfaVmZ/G7aRjEPJLL16TRZHAfdkEgWAu88zs78DR6NkIcNYMikxqnlI951JVMlEMmASyp2fk3y6e/vp6eunpzdus6+fnt74HKev6e6hJy63Yb7T3dtHT5/T09dPb8pLf6VobgzJpzmTmGIyaWowmho3JK6mBlufYDLTmxoaEtPC+6x1G4zmpg3rNjUmttGQ2HZcNmt/me1lXjcazQ0NWdur1aXGek0WuwI3FJj+GPC+IY5FZMRIJqqxtQ4m6usPSaMnJrGePqc7kXS6E8kmMz8zPXeZnj6na30SS84PCa03LtPbv2H5zu5eevvDPnv7PWeZmNAS76utwchJNBsSV3NDA7tuO47LP7Rnxfdbr8liPLC8wPRlQEehFczsZOBkgMmTJ1cvMhEZUo0NRmNDqMyvd+4eE4rT0x8SUm//hoTSM0Ci6emPCScmvd5Bllm/jVjqyiS5sIwzefyoqhxbvSYLKDzo0oDlL3e/ArgCQkeC1QpKRGQgZhYvKcEo6j+5pVGvN4EvJ5QucnVQuMQhIiJVVK/J4jFCvUWuXYDHhzgWEZERr16TxY3AvmY2LTPBzKYAB8R5IiIyhOo1WVwJPA/cYGZHm9lRhLujXgB+UsvARERGorpMFrGF9sHA08AvgWuAecDB7r66lrGJiIxEdXs3VOwD6thaxyEiInVashARkfqiZCEiIkWZb2z3lnXMzBYD88vYxBbAkgqFs6kYacc80o4XdMwjRTnHvIO7Tyw0Y1gmi3KZ2Vx3n17rOIbSSDvmkXa8oGMeKap1zLoMJSIiRSlZiIhIUUoWhV1R6wBqYKQd80g7XtAxjxRVOWbVWYiISFEqWYiISFFKFiIiUtSISRZmtr2ZXWdmK8xspZn9r5mVNKSembWZ2UVmtsjM1prZ3Wb2tmrHXK6NPWYzm25mV5jZk2bWaWYLzOwaM5s6FHGXo5zvOWc7Z5mZm9md1Yizkso9ZjPb2cx+Z2ZL4t/3U2Z2ejVjLkeZ/8uTzewX8W+608yeNrNvmVl7teMuh5ltZ2aXx3NPZ/zbnFLiuhU5f42IZGFmo4FZwE7A8cDHgNcDs0v8I/kZ8EngHOAIYBHwVzPboyoBV0CZx/xBwngilwGHA18G9gLmmtn2VQu6TBX4njPbmQacDbxajTgrqdxjNrPpwL1AK/AJ4F3AJVCfw7yVc7xx/q3A24CvAe8Gfgr8B/DzKoZdCTsC7ycM/nZHynUrc/5y92H/AE4H+oAdE9OmAr3AGUXW3Z0wxOuJiWlNwFPAjbU+tiod88QC03YA+oHzan1s1TjmnO38ldAV/hzgzlofVxW/5wbCQGO/r/VxDNHxHhr/lw/NmX5+XH90rY9vsO8q8foT8TimlLBexc5fI6JkARwF3OPuz2YmuPs84O/A0SWs2wP8JrFuL/Br4J1m1lr5cCtio4/Z3RcXmDYfWAxsW+E4K6mc7xkAM/swoRR1VlUirLxyjvkgwuiTl1Ytusor53hb4vPKnOmvERKnVSjGinP3/o1ctWLnr5GSLHYFHi0w/THCP0uxdee5e2eBdVsIxcN6VM4x5zGznYFJwBNlxlVNZR2zmXUA3wO+5O7LKhxbtZRzzAfG5zYzu8fMeszsVTO7zMxGVTTKyinneG8FngEuMLNdzGyMmR1MKK382MM4OsNNxc5fIyVZjCdc68u1DOgoY93M/HpUzjFnMbMm4MeEksXPyg+taso95osIA25dVcGYqq2cY94mPv8GuBk4BLiQcJnj2koFWGEbfbzuvo6QIDOX31YBfwP+CHy2smHWjYqdv+p28KMqKNT6sJRip5Wxbq1VKu7vA/sD73b3Qn949WSjjtnM3gp8HNjL44XdTcjGfs+ZH4tXu/s58fUcM2sEzjezXdz98YpEWFkb+x23ERLjJELF+AJgb0LFby/w6QrGWC8qdv4aKcliOYUzaAeFs27SMqDQbXkdifn1qJxjXs/MvgucDBzv7jdXKLZqKeeYf0IoNb1oZpvHaU1AY3y/1t27KhRnJZVzzEvj8y05028mVPruAdRbsijneP+dUE+zo7s/F6fdbmYrgCvM7Mfu/lDFIq0PFTt/jZTLUI8Rrt3l2oXi/wyPAVPjLXu563YDz+avUhfKOWYAzOxswm2zp7v7LysYW7WUc8w7A6cQTjiZxwHAvvF1vf7qLPdvG/J/eWZ+dW5spWo1lXO8bwaWJxJFxn3xeecyY6tHFTt/jZRkcSOwb7x/HoDYoOWAOK/Yus3A+xLrNgEfAG6u01+bUN4xY2anAd8Cznb3y6sVZIWVc8wzCzweIlSmzgSuq0K8lVDOMf8Z6AIOy5n+zvg8t0IxVlI5x/sy0GFmuZW6+8TnhZUKso5U7vxV6/uHh+ge5XZCBn2EcHvdUYQTwb+AMYnldiBcuzwnZ/1fE35dfgJ4O+HEsY5wfbvmx1fpYyY0yusnnEz2zXnsUutjq9b3XGB7c6j/dhbl/m1/PU7/DvAOQklyLXBVrY+t0scLTCHcNvs0oUHfTOCLcdpcEm0Z6vEBHBcfPyKUBj8d388o8h1X5PxV8w9gCD/oycD18Q9jFfB/5DRqiX9MDpybM30U4V70l+OHfC9wUK2PqVrHTLgbyAd4zKn1cVXrey6wrbpPFuUeM+GS0xnxBNxNGI74PKC51sdVpePdBfgt8AIhKT4NXAx01Pq4SjjuQf8nq33+UhflIiJS1EipsxARkTIoWYiISFFKFiIiUpSShYiIFKVkISIiRSlZiIhIUUoWmwgz28/MfmtmL5lZt5ktNbNbzOz42PFbtfZ7pJk9Ymbr4lCOm5vZHDObU8K655rZJntvtplNicd8Qq1jqSdmtkf8buu1x2WpgpHSkeAmzcw+R2hUMws4k9BwqoMw8tePCIO33FCF/TYB1wB3AZ8hNNpaBfy/Su+rTi0C9gNy+xIa6fYgtPy+mvrtSFMqTMmizsWB1S8Fvu/up+XMvsHMLiV0gVAN2wJjgd+6++2J6fXWE2lVeOg3555ax1EqM2v1An39DDRdKmPEfL61bsKuR9Em/n8ClgBtJS6/N2FEsNXAGsLgLnvnLHMV8CKwJ2Hw907CCGKnJJY5l4G7FZhDTrcfiW2tI3TI9jXgG+FPLGu5JsKQpU8SOrF7CbgkeXxs6LbgU4SuJxYRSk9/ALYrcMyfBB4kdN+wHLgN2D8xfzRwATCPUDqaB5xNkb6AEnGckOazK7LNicAPCd1NdMXnXwKtiWUOA+6Ox7OC0J3FG3O2Mwe4EzgS+Efc1ucJXXA78F7gSsKAVa/lfFYPxe9pCaFb9vEFvqMzCT8K1sVt/AXYCTihwN+FM8h40IS+xmbF7ayO8R5fYDkndF55WvyOVsXvctec5d5JKO2uiNt7itgfEjA9bufAxPKnZradmPb6OO1diWlTCSXpxfHz/Cfwnpx9nxvXexNhrPbVwA21Pk8MxaPmAegxyJcDjfFkdG2Jy+8WTzAPEDoYOxa4P07bPbHcVYR+dZ4gnJAPIYyM5sDMuMx2cRsOfJNEJ4LkJAtgC8JJ+glCb5bHEMZEfoH8ZPFrQhI7h9Bx3amERHB9Ypkpcb/Px7gOJ3T8tgS4LWd7F8dlf0o4cb47xvvBOL+JcFJfCnyO0JHa2YST4CVFPs9MHCek+ewG2V4HIbEsJZzY3w58KH4mY+MyhwF9hDEmjgI+TOi3aTGwbWJbc4BXCSfVkwhJYjc2JIuF8TM5DDgmrnM+YTzmSwiXME+My90LNCa2fR2hQ7qLM+sTSrczCcnum3Efx7Ghg8nWQY77K4RLl4fG7/y8GMcpOctlvvO/xmM/Lh7fs0BTXGYa4UR+TYzt4Pg9XBDnNxD+FpMdCP6e8H90V2Lap+IxZj737ePn+SjwUUJC+jmhQ82jEuudG+N8Lh7XwWwC/cRV5HxU6wD0GOTLgS3jH+Z3S1z+OsKJd/PEtM0I15X/NzHtKnJObkAr4WR8RWLajuScLOP0OWQni28TfrFPTkxrj9vzxLS3xu19PGd7H4nT94jvp8T3uYnhC3H6Non4+oBLB/lMPhbXeVvO9LNjzJMGWTcTxwmJaSV9dgNs77wY756DLDOXkFCaEtOmEk6ulyamzYknsj1y1j8oxvf7AsfSR36PpAfE5Y+J7w+O708bJMYT4jI7bsTfdAMhgV8JPJQzz+OxNyemZX6w7J/zfrNB9nEDMDuxv2WEBNlD7JmWkKDvSazzM0JCnpCzrVuAfybenxv3f3raY9/UH7obanh5G/BHd38tM8HdVxL6tJ+Rs2ynu89OLNdF+EctNKpWMfsR/vEWJLa3hnDZKOkwwgn6ejNryjwII7Nl4k+6Kef9I/E5E+M7CCeDKwaJ7TDCDQF3FdhnM+FXcVob+9kdCtzv7v8oNNPM2oG9gN+4e29i+/MIJbXc7/B5d//nAPv6fc77Qwif1TU5n8O9hJJS5rM/lHAyvLLIsZTMzF5vZr8ys4WEE3YPobvsNxZY/BZ370m8z/3O/xnX/7WZHWdmkwpsYzawXxxGdQ9gc8LY4l2EHywQkuqsxDqHES75rsj5fP4K7G5mm+XsI/fzHfaULOrbUsIlpB1KXH484fp+rpfJH8x+eYHluoC2kqPbYGvglQLTc6dNAloI13l7Eo9X4/wJOcvn3mmTqUTMxJhZ/sVBYptE+Px6ch6Z0dFy91mKjf3sJjB4rB2ELsMH+g5zb1UttNxA8zIn1WfJ/yw2Y8PnMAFY5u5rB9l2ycxsDOHX+e6EsTLeCryFcImntcAqg37n7v4s4RJRA6Gu52Uzu9fMkol0Vtz2/sRBrNz9FUIdz0wz25VQap+dWGcSYQz23M/mojg/9+9ksM9+WNLdUHXM3Xtje4ZDSrzjYhmwVYHpW1HdWxwXEf75cuVOW0qoK3hrgWUhVHansSQ+b0uo5CxkKeG69/sHmP98yn2WYwkh1oEsJ/yqH+g7XJozzQfZVu68zLqHUjjZZeYvAcab2agKJYz9CMn6re5+Z2Zi/NW+UWKpbraZtRIuo50H3GRmU9x9CaE0soRwSW1PNpQgZhH+Dl4glHD/ntjsUkLd1gUD7Db3b3Owz35YUsmi/p1P+FVzUaGZZjbVzHaLb28D3m1mYxPzxxIqfm+rYox3E4a63D6x3/a436S/EH4hjnP3uQUeaZPFrYTr9icPssxfCJWXqwfY55JB1q20m4G9zWz3QjPjpbsHgPclG1qa2Q6EX8nlfIe3ED6ryQN8DvMSMRrhMtFAMj9aRpWw38zYz+svLZlZB2GUu7K4e5e7zyJcYmon1O3goXLhNsKlt7eSnSz2BN4D3OvunYnN/YVwg8BjA3w+w//W2CJUsqhz7n67mZ0BXGpmOxMqWBcQLlm8nfBP/WHgYcJdKkcAfzOzCwi/fs4k/MOeV8Uwv0e42+VmMzuXcDL5IuESWvJY5pjZr4DrYvuQ+wgnsCnAu4Az3f3pUnfq7s+Z2feAM2JSvJFQibs38KS7/4Zw18yJhM/kEsJtoy3A6wh33ByTc9Kopu8RvqtbzexbhF/AWxBOnKe4+yrCLcc3AX80sx8CYwi3IK8gVNJulPhZXQB838zeSDiZriMk0kOAn7r7bHefbWbXE/7etiecYJsJdRo3ufscNrSz+YyZ/YKQCB529+4Cu76LUCfyAzP7OuGk/lXCL/9xaY/DzE6JsfyJUELYgnAr9kuEO5kyZgE/IPw93BGnPRhjmUn+/8M5hL/H283s+4QSZwfhFtlp7n5S2liHnVrXsOtR2oPwy/J3hEs+PYTLSjcTbvNrSCy3DyW2syiwjzlk3+VU0t1QcdpelNbOogE4nQ33+q+Iry8klDhgw11In8hZ96A4/aCc6acQkmVX/FzmAPsl5rcR7mJ5MrHM/XFaU+7nkFgvE8cJnvKzG2SbkwgV8osIl0JeAH7B4O0sbmCAdhYFtp/5jN4xwP4/RmhouCb+jTwBfJ9E+xXCj8izCUOOdhPuEvpTMgZCC+6FhJOxM3g7i4MJbSvWEm45PS1+9rl/G06iLUSh74BwWesGNrRTWUT4v8j9fHaO692TM/2GQn9Dcd52hNuNF8bjXkQokX00scy5cf0B/26G60PDqoqISFGqsxARkaKULEREpCglCxERKUrJQkREilKyEBGRopQsRESkKCULEREpSslCRESK+v8sFVuuJ7/CUgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "grid = np.linspace(0, 1, 1000)\n",
    "plt.plot(\n",
    "    grid,\n",
    "    [log_loss(y_true=np.array([1]), y_pred=np.array([g]), labels=(0, 1)) for g in grid],\n",
    ")\n",
    "plt.xlabel(\"Confidence in correct answer\")\n",
    "plt.ylabel(\"Loss (lower is better)\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Your loss goes to 0 as you approach 100% confidence in the correct answer.\n",
    "- Your loss goes to infinity as you approach 100% confidence in the incorrect answer.\n",
    "- (Optional) See also the very related [How to assign partial credit on an exam of true-false questions?](https://terrytao.wordpress.com/2016/06/01/how-to-assign-partial-credit-on-an-exam-of-true-false-questions/)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The real `LogisticRegression` is averaging this score over all training examples."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "\n",
    "Some nice examples: \n",
    "\n",
    "- [Scott Alexander](https://slatestarcodex.com/2019/01/22/2018-predictions-calibration-results/)\n",
    "  - Look at how the plot starts at 50%. That is because being 40% confident of \"X\" is the same as being 60% confident of \"not X\".\n",
    "- [Good Judgment Project](https://www.gjopen.com/)\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Credence Activity (time permitting: 15 min)\n",
    "\n",
    "- Take a few minutes and assign credences or values to the claims below, in the [Google Doc](https://docs.google.com/document/d/1y4YMSAWQGRFFn8ig0pAs5TJR5h2jKnvtY97HN-rEkyE/edit#). Afterwards, we'll discuss.\n",
    "- **Do not search the answers; the point of the exercise is to evaluate how good we are at guessing.** \n",
    "- Try not to be influenced by other peoples' answers! Better to pick your answers before going to the Google Doc.\n",
    "\n",
    "1. I am __ % sure that the world's longest river is between 6000km and 8000km. \n",
    "2. I am __ % sure that there is 4 to 8 liters of blood in an average adult human body. \n",
    "3. I am 99% sure that the world's tallest tree taller than __ m.\n",
    "4. I am 90% sure that the world's tallest tree taller than __ m.\n",
    "5. I am 50% sure that the world's tallest tree taller than __ m.\n",
    "\n",
    "> NOTE: 100% means you are completely sure the statement is true, 0% means you are completely sure the statement is false."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Visualizing your results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "- Very powerful but at the same time can be misleading if not done properly. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Pre-viewing review from [Calling BS visualization videos](https://www.youtube.com/watch?v=T-5aLbNeGo0&list=PLPnZfvKID1Sje5jWxt-4CSZD7bUI4gSPS&index=30&t=0s):\n",
    "\n",
    "- Dataviz in the popular media.\n",
    "  - e.g. [modern NYT](https://youtu.be/T-5aLbNeGo0?t=367)\n",
    "- Misleading axes.\n",
    "  - e.g. [vaccines](https://youtu.be/9pNWVMxaFuM?t=299)\n",
    "- Manipulating bin sizes.\n",
    "  - e.g. [tax dollars](https://youtu.be/zAg1wsYfwsM?t=196)\n",
    "- Dataviz ducks.\n",
    "  - e.g. [drinking water](https://youtu.be/rmii1hfP6d4?t=169)\n",
    "  - \"look how clever we are about design\" -> making it about me instead of about you (see last class)\n",
    "- Glass slippers.\n",
    "  - e.g. [internet marketing tree](https://youtu.be/59teS0SUHtI?t=285)\n",
    "- The principle of proportional ink.\n",
    "  - e.g. [most read books](https://youtu.be/oNhusd3xFC4?t=147)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- [Demo of cleaning up a plot](https://www.darkhorseanalytics.com/blog/data-looks-better-naked)\n",
    "- [Principle of proportional ink](https://serialmentor.com/dataviz/proportional-ink.html) from a viz textbook."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Dataset\n",
    "\n",
    "We'll be using [Kaggle House Prices dataset](https://www.kaggle.com/c/home-data-for-ml-course/), which we used in lecture 10. As usual, to run this notebook you'll need to download the data. For this dataset, train and test have already been separated. We'll be working with the train portion. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>Id</th>\n",
       "      <th>MSSubClass</th>\n",
       "      <th>MSZoning</th>\n",
       "      <th>LotFrontage</th>\n",
       "      <th>LotArea</th>\n",
       "      <th>Street</th>\n",
       "      <th>Alley</th>\n",
       "      <th>LotShape</th>\n",
       "      <th>LandContour</th>\n",
       "      <th>Utilities</th>\n",
       "      <th>...</th>\n",
       "      <th>PoolArea</th>\n",
       "      <th>PoolQC</th>\n",
       "      <th>Fence</th>\n",
       "      <th>MiscFeature</th>\n",
       "      <th>MiscVal</th>\n",
       "      <th>MoSold</th>\n",
       "      <th>YrSold</th>\n",
       "      <th>SaleType</th>\n",
       "      <th>SaleCondition</th>\n",
       "      <th>SalePrice</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>302</th>\n",
       "      <td>303</td>\n",
       "      <td>20</td>\n",
       "      <td>RL</td>\n",
       "      <td>118.0</td>\n",
       "      <td>13704</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>IR1</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2006</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "      <td>205000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>767</th>\n",
       "      <td>768</td>\n",
       "      <td>50</td>\n",
       "      <td>RL</td>\n",
       "      <td>75.0</td>\n",
       "      <td>12508</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>IR1</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Shed</td>\n",
       "      <td>1300</td>\n",
       "      <td>7</td>\n",
       "      <td>2008</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "      <td>160000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>429</th>\n",
       "      <td>430</td>\n",
       "      <td>20</td>\n",
       "      <td>RL</td>\n",
       "      <td>130.0</td>\n",
       "      <td>11457</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>IR1</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2009</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "      <td>175000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1139</th>\n",
       "      <td>1140</td>\n",
       "      <td>30</td>\n",
       "      <td>RL</td>\n",
       "      <td>98.0</td>\n",
       "      <td>8731</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>IR1</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2007</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "      <td>144000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>558</th>\n",
       "      <td>559</td>\n",
       "      <td>60</td>\n",
       "      <td>RL</td>\n",
       "      <td>57.0</td>\n",
       "      <td>21872</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>IR2</td>\n",
       "      <td>HLS</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>2008</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "      <td>175000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 81 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Id  MSSubClass MSZoning  LotFrontage  LotArea Street Alley LotShape  \\\n",
       "302    303          20       RL        118.0    13704   Pave   NaN      IR1   \n",
       "767    768          50       RL         75.0    12508   Pave   NaN      IR1   \n",
       "429    430          20       RL        130.0    11457   Pave   NaN      IR1   \n",
       "1139  1140          30       RL         98.0     8731   Pave   NaN      IR1   \n",
       "558    559          60       RL         57.0    21872   Pave   NaN      IR2   \n",
       "\n",
       "     LandContour Utilities  ... PoolArea PoolQC Fence MiscFeature MiscVal  \\\n",
       "302          Lvl    AllPub  ...        0    NaN   NaN         NaN       0   \n",
       "767          Lvl    AllPub  ...        0    NaN   NaN        Shed    1300   \n",
       "429          Lvl    AllPub  ...        0    NaN   NaN         NaN       0   \n",
       "1139         Lvl    AllPub  ...        0    NaN   NaN         NaN       0   \n",
       "558          HLS    AllPub  ...        0    NaN   NaN         NaN       0   \n",
       "\n",
       "     MoSold YrSold  SaleType  SaleCondition  SalePrice  \n",
       "302       1   2006        WD         Normal     205000  \n",
       "767       7   2008        WD         Normal     160000  \n",
       "429       3   2009        WD         Normal     175000  \n",
       "1139      5   2007        WD         Normal     144000  \n",
       "558       8   2008        WD         Normal     175000  \n",
       "\n",
       "[5 rows x 81 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"data/housing-kaggle/train.csv\")\n",
    "train_df, test_df = train_test_split(df, test_size=0.10, random_state=123)\n",
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1314, 81)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Let's separate `X` and `y`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "X_train = train_df.drop(columns=[\"SalePrice\"])\n",
    "y_train = train_df[\"SalePrice\"]\n",
    "\n",
    "X_test = test_df.drop(columns=[\"SalePrice\"])\n",
    "y_test = test_df[\"SalePrice\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Feature types \n",
    "\n",
    "- We have mixed feature types and a bunch of missing values. \n",
    "- Now, let's identify feature types and transformations. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "drop_features = [\"Id\"]\n",
    "numeric_features = [\n",
    "    \"BedroomAbvGr\",\n",
    "    \"KitchenAbvGr\",\n",
    "    \"LotFrontage\",\n",
    "    \"LotArea\",\n",
    "    \"OverallQual\",\n",
    "    \"OverallCond\",\n",
    "    \"YearBuilt\",\n",
    "    \"YearRemodAdd\",\n",
    "    \"MasVnrArea\",\n",
    "    \"BsmtFinSF1\",\n",
    "    \"BsmtFinSF2\",\n",
    "    \"BsmtUnfSF\",\n",
    "    \"TotalBsmtSF\",\n",
    "    \"1stFlrSF\",\n",
    "    \"2ndFlrSF\",\n",
    "    \"LowQualFinSF\",\n",
    "    \"GrLivArea\",\n",
    "    \"BsmtFullBath\",\n",
    "    \"BsmtHalfBath\",\n",
    "    \"FullBath\",\n",
    "    \"HalfBath\",\n",
    "    \"TotRmsAbvGrd\",\n",
    "    \"Fireplaces\",\n",
    "    \"GarageYrBlt\",\n",
    "    \"GarageCars\",\n",
    "    \"GarageArea\",\n",
    "    \"WoodDeckSF\",\n",
    "    \"OpenPorchSF\",\n",
    "    \"EnclosedPorch\",\n",
    "    \"3SsnPorch\",\n",
    "    \"ScreenPorch\",\n",
    "    \"PoolArea\",\n",
    "    \"MiscVal\",\n",
    "    \"YrSold\",\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['Po', 'Fa', 'TA', 'Gd', 'Ex'],\n",
       " ['Po', 'Fa', 'TA', 'Gd', 'Ex'],\n",
       " ['Po', 'Fa', 'TA', 'Gd', 'Ex'],\n",
       " ['Po', 'Fa', 'TA', 'Gd', 'Ex'],\n",
       " ['Po', 'Fa', 'TA', 'Gd', 'Ex'],\n",
       " ['Po', 'Fa', 'TA', 'Gd', 'Ex'],\n",
       " ['Po', 'Fa', 'TA', 'Gd', 'Ex'],\n",
       " ['Po', 'Fa', 'TA', 'Gd', 'Ex'],\n",
       " ['Po', 'Fa', 'TA', 'Gd', 'Ex'],\n",
       " ['Po', 'Fa', 'TA', 'Gd', 'Ex']]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ordinal_features_reg = [\n",
    "    \"ExterQual\",\n",
    "    \"ExterCond\",\n",
    "    \"BsmtQual\",\n",
    "    \"BsmtCond\",\n",
    "    \"HeatingQC\",\n",
    "    \"KitchenQual\",\n",
    "    \"FireplaceQu\",\n",
    "    \"GarageQual\",\n",
    "    \"GarageCond\",\n",
    "    \"PoolQC\",\n",
    "]\n",
    "ordering = [\n",
    "    \"Po\",\n",
    "    \"Fa\",\n",
    "    \"TA\",\n",
    "    \"Gd\",\n",
    "    \"Ex\",\n",
    "]  # if N/A it will just impute something, per below\n",
    "ordering_ordinal_reg = [ordering] * len(ordinal_features_reg)\n",
    "ordering_ordinal_reg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "ordinal_features_oth = [\n",
    "    \"BsmtExposure\",\n",
    "    \"BsmtFinType1\",\n",
    "    \"BsmtFinType2\",\n",
    "    \"Functional\",\n",
    "    \"Fence\",\n",
    "]\n",
    "ordering_ordinal_oth = [\n",
    "    [\"NA\", \"No\", \"Mn\", \"Av\", \"Gd\"],\n",
    "    [\"NA\", \"Unf\", \"LwQ\", \"Rec\", \"BLQ\", \"ALQ\", \"GLQ\"],\n",
    "    [\"NA\", \"Unf\", \"LwQ\", \"Rec\", \"BLQ\", \"ALQ\", \"GLQ\"],\n",
    "    [\"Sal\", \"Sev\", \"Maj2\", \"Maj1\", \"Mod\", \"Min2\", \"Min1\", \"Typ\"],\n",
    "    [\"NA\", \"MnWw\", \"GdWo\", \"MnPrv\", \"GdPrv\"],\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "The remaining features are categorical features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Alley',\n",
       " 'Neighborhood',\n",
       " 'Electrical',\n",
       " 'GarageType',\n",
       " 'MSSubClass',\n",
       " 'RoofMatl',\n",
       " 'Street',\n",
       " 'MSZoning',\n",
       " 'Condition1',\n",
       " 'Exterior1st',\n",
       " 'Foundation',\n",
       " 'GarageFinish',\n",
       " 'SaleType',\n",
       " 'CentralAir',\n",
       " 'MiscFeature',\n",
       " 'LandContour',\n",
       " 'SaleCondition',\n",
       " 'HouseStyle',\n",
       " 'PavedDrive',\n",
       " 'MasVnrType',\n",
       " 'LandSlope',\n",
       " 'Exterior2nd',\n",
       " 'BldgType',\n",
       " 'Condition2',\n",
       " 'MoSold',\n",
       " 'Heating',\n",
       " 'RoofStyle',\n",
       " 'LotShape',\n",
       " 'LotConfig',\n",
       " 'Utilities']"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "categorical_features = list(\n",
    "    set(X_train.columns)\n",
    "    - set(numeric_features)\n",
    "    - set(ordinal_features_reg)\n",
    "    - set(ordinal_features_oth)\n",
    "    - set(drop_features)\n",
    ")\n",
    "categorical_features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Applying feature transformations\n",
    "\n",
    "- Since we have mixed feature types, let's use `ColumnTransformer` to apply different transformations on different features types.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.compose import ColumnTransformer, make_column_transformer\n",
    "\n",
    "numeric_transformer = make_pipeline(SimpleImputer(strategy=\"median\"), StandardScaler())\n",
    "ordinal_transformer_reg = make_pipeline(\n",
    "    SimpleImputer(strategy=\"most_frequent\"),\n",
    "    OrdinalEncoder(categories=ordering_ordinal_reg),\n",
    ")\n",
    "\n",
    "ordinal_transformer_oth = make_pipeline(\n",
    "    SimpleImputer(strategy=\"most_frequent\"),\n",
    "    OrdinalEncoder(categories=ordering_ordinal_oth),\n",
    ")\n",
    "\n",
    "categorical_transformer = make_pipeline(\n",
    "    SimpleImputer(strategy=\"constant\", fill_value=\"missing\"),\n",
    "    OneHotEncoder(handle_unknown=\"ignore\", sparse=False),\n",
    ")\n",
    "\n",
    "preprocessor = make_column_transformer(\n",
    "    (\"drop\", drop_features),\n",
    "    (numeric_transformer, numeric_features),\n",
    "    (ordinal_transformer_reg, ordinal_features_reg),\n",
    "    (ordinal_transformer_oth, ordinal_features_oth),\n",
    "    (categorical_transformer, categorical_features),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Examining the preprocessed data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "preprocessor.fit(X_train)\n",
    "# Calling fit to examine all the transformers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "ohe_columns = list(\n",
    "    preprocessor.named_transformers_[\"pipeline-4\"]\n",
    "    .named_steps[\"onehotencoder\"]\n",
    "    .get_feature_names(categorical_features)\n",
    ")\n",
    "new_columns = (\n",
    "    numeric_features + ordinal_features_reg + ordinal_features_oth + ohe_columns\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "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>BedroomAbvGr</th>\n",
       "      <th>KitchenAbvGr</th>\n",
       "      <th>LotFrontage</th>\n",
       "      <th>LotArea</th>\n",
       "      <th>OverallQual</th>\n",
       "      <th>OverallCond</th>\n",
       "      <th>YearBuilt</th>\n",
       "      <th>YearRemodAdd</th>\n",
       "      <th>MasVnrArea</th>\n",
       "      <th>BsmtFinSF1</th>\n",
       "      <th>...</th>\n",
       "      <th>LotShape_IR2</th>\n",
       "      <th>LotShape_IR3</th>\n",
       "      <th>LotShape_Reg</th>\n",
       "      <th>LotConfig_Corner</th>\n",
       "      <th>LotConfig_CulDSac</th>\n",
       "      <th>LotConfig_FR2</th>\n",
       "      <th>LotConfig_FR3</th>\n",
       "      <th>LotConfig_Inside</th>\n",
       "      <th>Utilities_AllPub</th>\n",
       "      <th>Utilities_NoSeWa</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>302</th>\n",
       "      <td>0.154795</td>\n",
       "      <td>-0.222647</td>\n",
       "      <td>2.312501</td>\n",
       "      <td>0.381428</td>\n",
       "      <td>0.663680</td>\n",
       "      <td>-0.512408</td>\n",
       "      <td>0.993969</td>\n",
       "      <td>0.840492</td>\n",
       "      <td>0.269972</td>\n",
       "      <td>-0.961498</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>767</th>\n",
       "      <td>1.372763</td>\n",
       "      <td>-0.222647</td>\n",
       "      <td>0.260890</td>\n",
       "      <td>0.248457</td>\n",
       "      <td>-0.054669</td>\n",
       "      <td>1.285467</td>\n",
       "      <td>-1.026793</td>\n",
       "      <td>0.016525</td>\n",
       "      <td>-0.573129</td>\n",
       "      <td>0.476092</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>429</th>\n",
       "      <td>0.154795</td>\n",
       "      <td>-0.222647</td>\n",
       "      <td>2.885044</td>\n",
       "      <td>0.131607</td>\n",
       "      <td>-0.054669</td>\n",
       "      <td>-0.512408</td>\n",
       "      <td>0.563314</td>\n",
       "      <td>0.161931</td>\n",
       "      <td>-0.573129</td>\n",
       "      <td>1.227559</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1139</th>\n",
       "      <td>0.154795</td>\n",
       "      <td>-0.222647</td>\n",
       "      <td>1.358264</td>\n",
       "      <td>-0.171468</td>\n",
       "      <td>-0.773017</td>\n",
       "      <td>-0.512408</td>\n",
       "      <td>-1.689338</td>\n",
       "      <td>-1.679877</td>\n",
       "      <td>-0.573129</td>\n",
       "      <td>0.443419</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>558</th>\n",
       "      <td>0.154795</td>\n",
       "      <td>-0.222647</td>\n",
       "      <td>-0.597924</td>\n",
       "      <td>1.289541</td>\n",
       "      <td>0.663680</td>\n",
       "      <td>-0.512408</td>\n",
       "      <td>0.828332</td>\n",
       "      <td>0.598149</td>\n",
       "      <td>-0.573129</td>\n",
       "      <td>0.354114</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 263 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      BedroomAbvGr  KitchenAbvGr  LotFrontage   LotArea  OverallQual  \\\n",
       "302       0.154795     -0.222647     2.312501  0.381428     0.663680   \n",
       "767       1.372763     -0.222647     0.260890  0.248457    -0.054669   \n",
       "429       0.154795     -0.222647     2.885044  0.131607    -0.054669   \n",
       "1139      0.154795     -0.222647     1.358264 -0.171468    -0.773017   \n",
       "558       0.154795     -0.222647    -0.597924  1.289541     0.663680   \n",
       "\n",
       "      OverallCond  YearBuilt  YearRemodAdd  MasVnrArea  BsmtFinSF1  ...  \\\n",
       "302     -0.512408   0.993969      0.840492    0.269972   -0.961498  ...   \n",
       "767      1.285467  -1.026793      0.016525   -0.573129    0.476092  ...   \n",
       "429     -0.512408   0.563314      0.161931   -0.573129    1.227559  ...   \n",
       "1139    -0.512408  -1.689338     -1.679877   -0.573129    0.443419  ...   \n",
       "558     -0.512408   0.828332      0.598149   -0.573129    0.354114  ...   \n",
       "\n",
       "      LotShape_IR2  LotShape_IR3  LotShape_Reg  LotConfig_Corner  \\\n",
       "302            0.0           0.0           0.0               1.0   \n",
       "767            0.0           0.0           0.0               0.0   \n",
       "429            0.0           0.0           0.0               1.0   \n",
       "1139           0.0           0.0           0.0               0.0   \n",
       "558            1.0           0.0           0.0               0.0   \n",
       "\n",
       "      LotConfig_CulDSac  LotConfig_FR2  LotConfig_FR3  LotConfig_Inside  \\\n",
       "302                 0.0            0.0            0.0               0.0   \n",
       "767                 0.0            0.0            0.0               1.0   \n",
       "429                 0.0            0.0            0.0               0.0   \n",
       "1139                0.0            0.0            0.0               1.0   \n",
       "558                 0.0            1.0            0.0               0.0   \n",
       "\n",
       "      Utilities_AllPub  Utilities_NoSeWa  \n",
       "302                1.0               0.0  \n",
       "767                1.0               0.0  \n",
       "429                1.0               0.0  \n",
       "1139               1.0               0.0  \n",
       "558                1.0               0.0  \n",
       "\n",
       "[5 rows x 263 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_enc = pd.DataFrame(\n",
    "    preprocessor.transform(X_train), index=X_train.index, columns=new_columns\n",
    ")\n",
    "X_train_enc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>BedroomAbvGr</th>\n",
       "      <th>KitchenAbvGr</th>\n",
       "      <th>LotFrontage</th>\n",
       "      <th>LotArea</th>\n",
       "      <th>OverallQual</th>\n",
       "      <th>OverallCond</th>\n",
       "      <th>YearBuilt</th>\n",
       "      <th>YearRemodAdd</th>\n",
       "      <th>MasVnrArea</th>\n",
       "      <th>BsmtFinSF1</th>\n",
       "      <th>...</th>\n",
       "      <th>LotShape_IR2</th>\n",
       "      <th>LotShape_IR3</th>\n",
       "      <th>LotShape_Reg</th>\n",
       "      <th>LotConfig_Corner</th>\n",
       "      <th>LotConfig_CulDSac</th>\n",
       "      <th>LotConfig_FR2</th>\n",
       "      <th>LotConfig_FR3</th>\n",
       "      <th>LotConfig_Inside</th>\n",
       "      <th>Utilities_AllPub</th>\n",
       "      <th>Utilities_NoSeWa</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>0.154795</td>\n",
       "      <td>-0.222647</td>\n",
       "      <td>-0.025381</td>\n",
       "      <td>-0.085415</td>\n",
       "      <td>0.663680</td>\n",
       "      <td>-0.512408</td>\n",
       "      <td>0.993969</td>\n",
       "      <td>0.792023</td>\n",
       "      <td>0.438592</td>\n",
       "      <td>-0.961498</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>676</th>\n",
       "      <td>1.372763</td>\n",
       "      <td>4.348569</td>\n",
       "      <td>-0.454788</td>\n",
       "      <td>-0.074853</td>\n",
       "      <td>-1.491366</td>\n",
       "      <td>-3.209221</td>\n",
       "      <td>-2.351883</td>\n",
       "      <td>-1.679877</td>\n",
       "      <td>-0.573129</td>\n",
       "      <td>-0.961498</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1304</th>\n",
       "      <td>0.154795</td>\n",
       "      <td>-0.222647</td>\n",
       "      <td>-1.790721</td>\n",
       "      <td>-0.768279</td>\n",
       "      <td>0.663680</td>\n",
       "      <td>-0.512408</td>\n",
       "      <td>1.093350</td>\n",
       "      <td>0.937429</td>\n",
       "      <td>0.084490</td>\n",
       "      <td>-0.961498</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1372</th>\n",
       "      <td>0.154795</td>\n",
       "      <td>-0.222647</td>\n",
       "      <td>0.260890</td>\n",
       "      <td>-0.058176</td>\n",
       "      <td>0.663680</td>\n",
       "      <td>0.386530</td>\n",
       "      <td>0.894587</td>\n",
       "      <td>0.646617</td>\n",
       "      <td>-0.573129</td>\n",
       "      <td>1.162214</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1427</th>\n",
       "      <td>0.154795</td>\n",
       "      <td>-0.222647</td>\n",
       "      <td>-0.454788</td>\n",
       "      <td>0.073016</td>\n",
       "      <td>-0.773017</td>\n",
       "      <td>0.386530</td>\n",
       "      <td>-0.861157</td>\n",
       "      <td>-1.679877</td>\n",
       "      <td>-0.573129</td>\n",
       "      <td>0.301838</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 263 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      BedroomAbvGr  KitchenAbvGr  LotFrontage   LotArea  OverallQual  \\\n",
       "147       0.154795     -0.222647    -0.025381 -0.085415     0.663680   \n",
       "676       1.372763      4.348569    -0.454788 -0.074853    -1.491366   \n",
       "1304      0.154795     -0.222647    -1.790721 -0.768279     0.663680   \n",
       "1372      0.154795     -0.222647     0.260890 -0.058176     0.663680   \n",
       "1427      0.154795     -0.222647    -0.454788  0.073016    -0.773017   \n",
       "\n",
       "      OverallCond  YearBuilt  YearRemodAdd  MasVnrArea  BsmtFinSF1  ...  \\\n",
       "147     -0.512408   0.993969      0.792023    0.438592   -0.961498  ...   \n",
       "676     -3.209221  -2.351883     -1.679877   -0.573129   -0.961498  ...   \n",
       "1304    -0.512408   1.093350      0.937429    0.084490   -0.961498  ...   \n",
       "1372     0.386530   0.894587      0.646617   -0.573129    1.162214  ...   \n",
       "1427     0.386530  -0.861157     -1.679877   -0.573129    0.301838  ...   \n",
       "\n",
       "      LotShape_IR2  LotShape_IR3  LotShape_Reg  LotConfig_Corner  \\\n",
       "147            0.0           0.0           0.0               0.0   \n",
       "676            0.0           0.0           1.0               0.0   \n",
       "1304           0.0           0.0           1.0               0.0   \n",
       "1372           0.0           0.0           1.0               1.0   \n",
       "1427           0.0           0.0           1.0               0.0   \n",
       "\n",
       "      LotConfig_CulDSac  LotConfig_FR2  LotConfig_FR3  LotConfig_Inside  \\\n",
       "147                 1.0            0.0            0.0               0.0   \n",
       "676                 0.0            0.0            0.0               1.0   \n",
       "1304                0.0            0.0            0.0               1.0   \n",
       "1372                0.0            0.0            0.0               0.0   \n",
       "1427                0.0            0.0            0.0               1.0   \n",
       "\n",
       "      Utilities_AllPub  Utilities_NoSeWa  \n",
       "147                1.0               0.0  \n",
       "676                1.0               0.0  \n",
       "1304               1.0               0.0  \n",
       "1372               1.0               0.0  \n",
       "1427               1.0               0.0  \n",
       "\n",
       "[5 rows x 263 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_enc = pd.DataFrame(\n",
    "    preprocessor.transform(X_test), index=X_test.index, columns=new_columns\n",
    ")\n",
    "X_test_enc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((1314, 80), (146, 80))"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Training random forests and gradient boosted trees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingRegressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's compare sklearn's `GradientBoostingRegressor` to `RandomForestRegressor` for different values of `n_estimators`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_estimators_values = [3, 10, 30, 100, 300]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3\n",
      "10\n",
      "30\n",
      "100\n",
      "300\n"
     ]
    }
   ],
   "source": [
    "score_rf_train = list()\n",
    "score_rf_test = list()\n",
    "score_gb_train = list()\n",
    "score_gb_test = list()\n",
    "\n",
    "for n_estimators in n_estimators_values:\n",
    "    print(n_estimators)\n",
    "    rf = TransformedTargetRegressor(\n",
    "        RandomForestRegressor(n_estimators=n_estimators, random_state=123),\n",
    "        func=np.log1p,\n",
    "        inverse_func=np.expm1,\n",
    "    )\n",
    "    rf.fit(X_train_enc, y_train)\n",
    "    score_rf_train.append(rf.score(X_train_enc, y_train))\n",
    "    score_rf_test.append(rf.score(X_test_enc, y_test))\n",
    "\n",
    "    gb = TransformedTargetRegressor(\n",
    "        GradientBoostingRegressor(n_estimators=n_estimators, random_state=123),\n",
    "        func=np.log1p,\n",
    "        inverse_func=np.expm1,\n",
    "    )\n",
    "    gb.fit(X_train_enc, y_train)\n",
    "    score_gb_train.append(gb.score(X_train_enc, y_train))\n",
    "    score_gb_test.append(gb.score(X_test_enc, y_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Here is a low-quality plot that is confusing and perhaps downright misleading:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.subplot(2, 1, 1)\n",
    "plt.plot(n_estimators_values, score_rf_train, label=\"rf train\")\n",
    "plt.plot(n_estimators_values, score_rf_test, label=\"rf test\")\n",
    "plt.ylabel(\"$R^2$ score\")\n",
    "plt.legend()\n",
    "plt.subplot(2, 1, 2)\n",
    "plt.plot(n_estimators_values, score_gb_train, label=\"gb train\")\n",
    "plt.plot(n_estimators_values, score_gb_test, label=\"gb test\")\n",
    "plt.xlabel(\"n estimators\")\n",
    "plt.ylabel(\"$R^2$ score\")\n",
    "plt.legend()\n",
    "plt.tight_layout();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Let's create some visualizations. \n",
    "\n",
    "- Create a visualization that makes RF look better than GB.\n",
    "- Create a visualization that makes GB look better than RF.\n",
    "- Create a visualization that makes RF and GB look equally good.\n",
    "\n",
    "You can create your own misleading example and copy or screenshot it and paste it into the [Google Doc](https://docs.google.com/document/d/1y4YMSAWQGRFFn8ig0pAs5TJR5h2jKnvtY97HN-rEkyE/edit#).\n",
    "\n",
    "Here are some misleading plots. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### RF better than GB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "nmax = 3\n",
    "\n",
    "# plt.plot(n_estimators_values[:nmax], score_rf_train[:nmax], label=\"rf train\")\n",
    "plt.plot(n_estimators_values[:nmax], score_rf_test[:nmax], label=\"rf test\")\n",
    "# plt.plot(n_estimators_values[:nmax], score_gb_train[:nmax], label=\"gb train\")\n",
    "plt.plot(n_estimators_values[:nmax], score_gb_test[:nmax], label=\"gb test\")\n",
    "plt.xlabel(\"n estimators\")\n",
    "plt.ylabel(\"$R^2$ score\")\n",
    "plt.legend()\n",
    "plt.title(\"for most values of n_estimators, RF is betteR!!\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### GB better than RF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plt.plot(n_estimators_values, score_rf_train, label=\"rf train\")\n",
    "plt.plot(n_estimators_values, score_rf_test, label=\"rf test\")\n",
    "# plt.ylabel(\"$R^2$ score\");\n",
    "# plt.legend();\n",
    "# plt.subplot(2,1,2)\n",
    "# plt.plot(n_estimators_values, score_gb_train, label=\"gb train\")\n",
    "plt.plot(n_estimators_values, score_gb_test, label=\"gb test\")\n",
    "plt.xlabel(\"n estimators\")\n",
    "plt.ylabel(\"$R^2$ score\")\n",
    "plt.legend()\n",
    "plt.title(\"GB better!!!\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Equally good"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "nmax = 2\n",
    "\n",
    "# plt.plot(n_estimators_values, score_rf_train, label=\"rf train\")\n",
    "plt.plot(n_estimators_values[:nmax], score_rf_test[:nmax], \"b\", label=\"rf test\")\n",
    "plt.ylabel(\"RF $R^2$ score\")\n",
    "plt.legend(loc=\"lower left\")\n",
    "plt.ylim((0.8, 0.9))\n",
    "plt.twinx()\n",
    "# plt.plot(n_estimators_values, score_gb_train, label=\"gb train\")\n",
    "plt.plot(n_estimators_values[:nmax], score_gb_test[:nmax], \"--r\", label=\"gb test\")\n",
    "plt.xlabel(\"n estimators\")\n",
    "plt.ylabel(\"GB $R^2$ score\")\n",
    "plt.legend()\n",
    "plt.ylim((-0.01, 0.70))\n",
    "plt.title(\"Both equally good!!!\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "**Be critical of your visualizations and try to make them as honest as possible.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## What did we learn today?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Principles of effective communication\n",
    "  - Concepts then labels, not the other way around.\n",
    "  - Bottom-up explanations.\n",
    "  - New ideas in small chunks.\n",
    "  - Reuse your running examples.\n",
    "  - Approaches from all angles.  \n",
    "  - When experimenting, show the results asap.\n",
    "  - **It's not about you.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- Decision variables, objectives, and context.\n",
    "- How does ML fit in?\n",
    "- Expressing your confidence about the results\n",
    "- Misleading visualizations.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Have a great weekend! \n",
    "\n",
    "![](img/eva-seeyou.png)"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "kernelspec": {
   "display_name": "Python [conda env:cpsc330]",
   "language": "python",
   "name": "conda-env-cpsc330-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}