Bootstrap.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  Chọn mẫu ngẫu nhiên lặp lại\n",
    "\n",
    "## BS. Lê Ngọc Khả Nhi\n",
    "\n",
    "Đây là bài thực hành đầu tiên nằm trong Series lập trình thống kê Python nâng cao. Gọi là nâng cao vì nó đòi hỏi một số kỹ năng viết code nhất định, nhưng nội dung sẽ không quá phức tạp. Trong bài này, Nhi sẽ hướng dẫn các bạn thực hiện một quy trình chọn mẫu ngẫu nhiên lặp lại (hay Bootstrap). Mục tiêu của bootstrap là ước lượng phân bố của một hay nhiều trị số thống kê từ dữ liệu.\n",
    "\n",
    "Bạn có thể làm bootstrap cho mọi thứ, từ các chỉ số thống kê mô tả (trung bình, trung vị) cho đến  hệ số tương quan Pearson, Odds-ratio, điểm số phẩm chất mô hình (Recall, Accuracy, F1 score, Kappa...), AUC của ROC curve...\n",
    "\n",
    "Một quy trình bootstrap thường có cấu trúc như sau:\n",
    "\n",
    "(1) Đầu tiên, ta cần chuẩn bị một place-holder để thu nhận kết quả sinh ra từ vòng lặp tái chọn mẫu. Placeholder là 1 data object, có cấu trúc tùy theo nhu cầu về kết quả mà ta mong muốn. Ta có thể dùng list, dictionary, numpy array hoặc pandas dataframes.\n",
    "\n",
    "(2) Pseudo-random value generator: Một cơ chế để sinh ra giá trị ngẫu nhiên; thông thường ta dùng các method random của python hay của module numpy.random, thí dụ randint hay random.choice, để thực hiện việc chọn mẫu ngẫu nhiên.\n",
    "\n",
    "(3) Quy trình (hàm) để ước tính một hay nhiều giá trị mà ta mong muốn trên mẫu dữ liệu sinh ra từ (2) \n",
    "\n",
    "(4) Một vòng lặp, để lặp lại N lần bước (2) và (3), và mỗi lần như vậy gán dữ liệu từ (3) vào place holder (1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from scipy import integrate\n",
    "from sklearn.metrics import auc\n",
    "\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import arviz as az"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thí dụ, ta có một bài toán giả định như sau:\n",
    "\n",
    "Cho spo2 là chuỗi giá trị của độ bão hòa oxy máu (SPO2) ghi nhận trong 180 phút với tần số lấy mẫu 1 lần/phút\n",
    "\n",
    "Mục tiêu: Ước tính giá trị trung bình của SPO2 trong mỗi 10 phút"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([94, 96, 95, 96, 94, 96, 95, 92, 96, 96, 95, 96, 96, 92, 95, 94, 95,\n",
       "       92, 93, 92, 93, 95, 95, 93, 95, 93, 95, 96, 94, 94, 96, 95, 94, 93,\n",
       "       96, 93, 95, 96, 94, 93, 94, 92, 93, 92, 95, 94, 94, 93, 94, 92, 95,\n",
       "       92, 95, 96, 95, 94, 92, 93, 93, 96, 94, 94, 92, 92, 95, 94, 92, 94,\n",
       "       96, 92, 94, 96, 95, 96, 92, 93, 92, 95, 92, 94, 92, 92, 94, 92, 93,\n",
       "       94, 95, 96, 94, 94, 93, 92, 92, 93, 94, 96, 92, 94, 96, 95, 94, 96,\n",
       "       92, 94, 95, 93, 95, 96, 94, 95, 92, 95, 93, 94, 96, 94, 94, 92, 92,\n",
       "       96, 94, 92, 93, 93, 95, 92, 92, 95, 92, 92, 94, 92, 96, 96, 95, 92,\n",
       "       94, 93, 93, 94, 96, 93, 92, 93, 94, 93, 93, 92, 92, 94, 94, 96, 96,\n",
       "       92, 92, 93, 92, 96, 95, 93, 94, 92, 96, 92, 93, 94, 94, 92, 95, 95,\n",
       "       96, 95, 93, 95, 96, 96, 94, 92, 96, 95])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spo2 = np.random.randint(low = 92, high = 97, size = 180)\n",
    "\n",
    "spo2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1fe43dc6388>]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1728x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "matplotlib.rcParams.update({'font.size': 20})\n",
    "plt.figure(figsize=(24,5))\n",
    "\n",
    "plt.fill_between(np.arange(spo2.shape[0]),\n",
    "                 np.repeat(90, spo2.shape[0]),\n",
    "                 spo2, \n",
    "                 step=\"pre\", \n",
    "                 alpha=0.3, \n",
    "                 color = 'red')\n",
    "plt.xlabel('Thời gian (phút)')\n",
    "plt.ylabel('SPO2')\n",
    "\n",
    "plt.plot(np.arange(spo2.shape[0]),spo2, drawstyle=\"steps\", color = 'red', alpha = 0.4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Trong không gian chọn mẫu spo2, mỗi  đơn vị quan sát tương ứng với 1 phút. Nếu ta không quan tâm đến trình tự thời gian, chỉ cần ta chọn ngẫu nhiên 10 đơn vị quan sát rồi lấy trung bình, và lặp lại rất nhiều lần, ta có thể trả lời được câu hỏi mục tiêu.\n",
    "\n",
    "Nhi viết 1 hàm boot_mean để thực hiện quy trình này:\n",
    "\n",
    "1) sampl là place holder, nó là 1 numpy array rỗng\n",
    "\n",
    "2) Cơ chế chọn mẫu ngẫu nhiên là method np.random.choice với size = 10 (10 phút)\n",
    "\n",
    "3) Giá trị cần tính là trung bình, dùng method np.mean\n",
    "\n",
    "4) Vòng lặp for lặp lại N lần"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def boot_mean(seq = None, n_iter = 100, window_size = 10):\n",
    "    \n",
    "    sampl = np.array([]) # place holder\n",
    "    \n",
    "    for i in range(n_iter):\n",
    "        s_mean = np.mean(np.random.choice(seq, replace = True, \n",
    "                                          size = window_size)\n",
    "                        )\n",
    "        sampl = np.append(sampl, s_mean)\n",
    "    \n",
    "    return sampl"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thi hành hàm boot_mean với số lượt lấy mẫu là 10 ngàn, và lưu kết quả vào array mean_spo2:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_spo2 = boot_mean(spo2, 10000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "10 giá trị đầu tiên của kết quả thu được:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([94.8, 93.4, 93.5, 93.6, 94.2, 94.1, 95.1, 93.5, 94. , 94. ])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_spo2[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Bây giờ, ta có thể dùng thống kê mô tả để khảo sát 10000 kết quả thu được:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, '')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "az.plot_posterior(mean_spo2,\n",
    "                  credible_interval = 0.95,\n",
    "                  point_estimate = 'median',\n",
    "                  kind = 'kde',\n",
    "                  color =\"red\")\n",
    "\n",
    "plt.xlabel('SPO2 trung bình / 10 phút')\n",
    "plt.title(None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Tuy nhiên, một số bạn có thể không hài lòng với cách làm này, với lập luận rằng: Spo2 là 1 chuỗi thời gian nên khi chọn mẫu ta cần đảm bảo tính liên tục theo trình tự thời gian, tức là ta phải chọn ngẫu nhiên những phân đoạn 10 giá trị liên tục chứ không phải lấy 10 điểm rải rác tại các thời điểm xa nhau.\n",
    "\n",
    "Ta có thể làm việc này 1 cách dễ dàng như sau:\n",
    "\n",
    "Dùng method numpy.random.randint để sinh ra ngẫu nhiên 10000 con số trong khoảng từ 0 đến 180, xem đây là điểm khởi đầu của phân đoạn ta cần lấy mẫu, cộng thêm 10 là ta có được điểm kết thúc. Dùng 1 cặp khởi đầu và kết thúc như indices array, ta có thể trích ra được phân đoạn 10 phút liên tục.\n",
    "\n",
    "Có một vấn đề nhỏ: nếu điểm kết thúc vượt quá giới hạn 180 thì sao ? Ta chỉ cần sửa bằng cách lùi điểm khởi đầu lại 10 vị trí, như vậy ta không bao giờ lấy mẫu vượt quá giới hạn độ dài của chuỗi spo2.\n",
    "\n",
    "Tiếp theo, ta có thể làm phức tạp hơn 1 chút: nếu ta muốn khảo sát kết quả trên toàn bộ thang đo của SpO2 (bao gồm giá trị min, max và toàn bộ các giá trị có thể giữa 2 mức này của spO2) nhưng kết quả lại không chứa các giá trị này thì sao ? Ta chỉ cần chuẩn bị place holder chứa sẵn thang đo cơ bản này, lưu ý là mỗi mốc trên thang đo chỉ xuất hiện với tần suất 1 lần, nên không có ảnh hưởng gì nhiều trên phân phối 10000 quan sát.\n",
    "\n",
    "Ta lại phức tạp vấn đề thêm 1 chút nữa: Nếu như ta không chỉ đơn giản muốn tính trung bình cộng của 10 giá trị như trên, nhưng muốn dùng trung bình tích phân (tức là tính diện tích dưới đường cong SpO2 trong phân đoạn t=10 với dt = 1, rồi chia cho t-1)\n",
    "\n",
    "Ta xây dựng 1 hàm cho phép làm tất cả việc trên:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def boot_auc(seq = None, n_iter = 100, window_size = 10, full_scale = True):\n",
    "    \n",
    "    def sample_sequence(seq = None, n_iter = 100, window_size = 10):\n",
    "        np.random.seed(123)\n",
    "        start_loc = np.random.randint(0, seq.shape[0], size = n_iter)\n",
    "        start_loc -= window_size*(start_loc > (seq.shape[0] - window_size))\n",
    "        return start_loc\n",
    "    \n",
    "    if full_scale:\n",
    "        auc_array = np.linspace(np.min(seq), np.max(seq), 10)\n",
    "    else:\n",
    "        auc_array = np.array([])\n",
    "    \n",
    "    locs = sample_sequence(seq, n_iter, window_size)\n",
    "    \n",
    "    for i in locs:\n",
    "        auc_array = np.append(auc_array,\n",
    "                              np.trapz(seq[i:i+10], dx = 1)/(window_size - 1))\n",
    "    return auc_array"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hàm này xuất ra chuỗi giá trị trung bình tích phân SpO2 của các phân đoạn 10 phút:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "auc_seqs = boot_auc(seq = spo2, \n",
    "                    n_iter = 10000, \n",
    "                    window_size = 10, \n",
    "                    full_scale = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, '')"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "az.plot_posterior(auc_seqs,\n",
    "                  credible_interval = 0.95,\n",
    "                  point_estimate = 'median',\n",
    "                  kind = 'kde',\n",
    "                  color =\"red\")\n",
    "\n",
    "plt.xlabel('SPO2')\n",
    "plt.title(None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Bài thực hành đến đây tạm dừng, các bạn đã hình dung ra được cấu trúc Python code của một quy trình bootstrap; dựa vào cấu trúc này các bạn có thể áp dụng bootstrap cho mọi vấn đề khác từ đơn giản đến phức tạp. Thí dụ nếu cần xuất ra nhiều giá trị, chỉ cần tạo place holder là dictionary , array đa chiều hay pandas dataframe. Nếu giá trị cần tính là phức tạp, chỉ cần viết hàm cho từng công đoạn nhỏ rồi ghép lại trong vòng lặp.\n",
    "\n",
    "Chúc các bạn thực hành vui và vận dụng thành công Python cho vấn đề của mình."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "hide_input": false,
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.4"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}