narrowlyapplicable/simple-example of lifetime.ipynb

## simple-example of lifetime.ipynb
{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 打ち切りを含むデータから製品寿命を推定する"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "M. Hamada,  *et al.* \"Bayesian Reliability\" の寿命推定例を実装した。\n",
    "- 打ち切りを含む製品寿命データから、真の寿命を推定する。  \n",
    "\n",
    "想定している状況\n",
    "- 複数の部品からなる装置を想定し、そのうち特定の部品の寿命を知りたい。\n",
    "  - 社内テストではなく、実際の使用環境下での寿命を調べたい。\n",
    "  - そのため、客先の装置における実例から部品寿命を推定することを考える。\n",
    "- 他の部品が故障すると、対象装置の寿命前に交換されてしまう（=右側打ち切り）\n",
    "  - しかも、打ち切りの期間＝他部品の寿命 であるため変動する。  \n",
    "  \n",
    "目的\n",
    "- 打ち切り期間が変動するデータから生存時間を推定する。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. ライブラリのインポート"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from pystan import StanModel\n",
    "import pickle\n",
    "plt.style.use('ggplot')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. データの読込・作成"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\"Bayesian Reliability\"[1]に掲載されている軸受の寿命データを用いる"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('./bearing.csv', header=None)\n",
    "df.rename(columns = {0:\"value\", 1:\"trunc\"}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "      <th>17</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>value</th>\n",
       "      <td>1085.0</td>\n",
       "      <td>646.0</td>\n",
       "      <td>1795.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>727.0</td>\n",
       "      <td>2238.0</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>820.0</td>\n",
       "      <td>1890.0</td>\n",
       "      <td>1628.0</td>\n",
       "      <td>2294.0</td>\n",
       "      <td>1388.0</td>\n",
       "      <td>1390.0</td>\n",
       "      <td>897.0</td>\n",
       "      <td>1145.0</td>\n",
       "      <td>759.0</td>\n",
       "      <td>663.0</td>\n",
       "      <td>1153.0</td>\n",
       "      <td>152.0</td>\n",
       "      <td>810.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>trunc</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           0      1       2      3      4       5       6      7       8   \\\n",
       "value  1085.0  646.0  1795.0  100.0  727.0  2238.0  1500.0  820.0  1890.0   \n",
       "trunc     1.0    NaN     1.0    1.0    1.0     1.0     1.0    1.0     NaN   \n",
       "\n",
       "           9       10      11      12     13      14     15     16      17  \\\n",
       "value  1628.0  2294.0  1388.0  1390.0  897.0  1145.0  759.0  663.0  1153.0   \n",
       "trunc     NaN     1.0     1.0     1.0    NaN     1.0    1.0    1.0     1.0   \n",
       "\n",
       "          18     19  \n",
       "value  152.0  810.0  \n",
       "trunc    1.0    1.0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[:20].T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>20</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "      <th>31</th>\n",
       "      <th>32</th>\n",
       "      <th>33</th>\n",
       "      <th>34</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>38</th>\n",
       "      <th>39</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>value</th>\n",
       "      <td>1380.0</td>\n",
       "      <td>246.0</td>\n",
       "      <td>1427.0</td>\n",
       "      <td>2892.0</td>\n",
       "      <td>971.0</td>\n",
       "      <td>80.0</td>\n",
       "      <td>61.0</td>\n",
       "      <td>861.0</td>\n",
       "      <td>951.0</td>\n",
       "      <td>2153.0</td>\n",
       "      <td>966.0</td>\n",
       "      <td>1167.0</td>\n",
       "      <td>1165.0</td>\n",
       "      <td>462.0</td>\n",
       "      <td>767.0</td>\n",
       "      <td>853.0</td>\n",
       "      <td>997.0</td>\n",
       "      <td>711.0</td>\n",
       "      <td>437.0</td>\n",
       "      <td>1079.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>trunc</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           20     21      22      23     24    25    26     27     28      29  \\\n",
       "value  1380.0  246.0  1427.0  2892.0  971.0  80.0  61.0  861.0  951.0  2153.0   \n",
       "trunc     1.0    1.0     1.0     1.0    1.0   1.0   1.0    1.0    NaN     1.0   \n",
       "\n",
       "          30      31      32     33     34     35     36     37     38      39  \n",
       "value  966.0  1167.0  1165.0  462.0  767.0  853.0  997.0  711.0  437.0  1079.0  \n",
       "trunc    1.0     NaN     1.0    1.0    1.0    1.0    1.0    NaN    1.0     1.0  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[20:40].T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>40</th>\n",
       "      <th>41</th>\n",
       "      <th>42</th>\n",
       "      <th>43</th>\n",
       "      <th>44</th>\n",
       "      <th>45</th>\n",
       "      <th>46</th>\n",
       "      <th>47</th>\n",
       "      <th>48</th>\n",
       "      <th>49</th>\n",
       "      <th>50</th>\n",
       "      <th>51</th>\n",
       "      <th>52</th>\n",
       "      <th>53</th>\n",
       "      <th>54</th>\n",
       "      <th>55</th>\n",
       "      <th>56</th>\n",
       "      <th>57</th>\n",
       "      <th>58</th>\n",
       "      <th>59</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>value</th>\n",
       "      <td>546.0</td>\n",
       "      <td>911.0</td>\n",
       "      <td>887.0</td>\n",
       "      <td>1203.0</td>\n",
       "      <td>1152.0</td>\n",
       "      <td>1199.0</td>\n",
       "      <td>736.0</td>\n",
       "      <td>2181.0</td>\n",
       "      <td>977.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>159.0</td>\n",
       "      <td>424.0</td>\n",
       "      <td>917.0</td>\n",
       "      <td>1042.0</td>\n",
       "      <td>1022.0</td>\n",
       "      <td>1070.0</td>\n",
       "      <td>3428.0</td>\n",
       "      <td>763.0</td>\n",
       "      <td>2871.0</td>\n",
       "      <td>799.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>trunc</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          40     41     42      43      44      45     46      47     48  \\\n",
       "value  546.0  911.0  887.0  1203.0  1152.0  1199.0  736.0  2181.0  977.0   \n",
       "trunc    1.0    1.0    1.0     NaN     1.0     1.0    1.0     NaN    1.0   \n",
       "\n",
       "         49     50     51     52      53      54      55      56     57  \\\n",
       "value  85.0  159.0  424.0  917.0  1042.0  1022.0  1070.0  3428.0  763.0   \n",
       "trunc   1.0    1.0    1.0    1.0     1.0     1.0     1.0     1.0    1.0   \n",
       "\n",
       "           58     59  \n",
       "value  2871.0  799.0  \n",
       "trunc     1.0    1.0  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[40:60].T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>60</th>\n",
       "      <th>61</th>\n",
       "      <th>62</th>\n",
       "      <th>63</th>\n",
       "      <th>64</th>\n",
       "      <th>65</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>value</th>\n",
       "      <td>2087.0</td>\n",
       "      <td>719.0</td>\n",
       "      <td>555.0</td>\n",
       "      <td>1297.0</td>\n",
       "      <td>1231.0</td>\n",
       "      <td>750.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>trunc</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           60     61     62      63      64     65\n",
       "value  2087.0  719.0  555.0  1297.0  1231.0  750.0\n",
       "trunc     1.0    1.0    1.0     1.0     1.0    NaN"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[60:].T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- １列目には、軸受の取付〜交換までの稼働時間が記録されている。\n",
    "- ２列目には、打ち切り（この場合であれば、交換前に故障が生じていたか）の有無が記録されている。故障により交換された場合には何も記入されないが、交換時に未だ故障していなかった場合、２列目に「１」が記録される。\n",
    "  - 例）index=0の軸受は1085時間が経過した時点で交換されたが、交換された時点では故障していなかった。\n",
    "  - 例）index=1の軸受は646時間で破損し、すぐに交換された。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. モデルの定義"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "原著で使用されたモデルは以下の通り"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- モデル\n",
    "$$f(t_u, t_{tc} | \\mu, \\sigma^2) = \\prod_{i=1}^Nf(t_{u,i} | \\mu, \\sigma^2) \\times \\prod_{i=1}^N(1-F(t_{tc, i}|\\mu, \\sigma^2))$$\n",
    "  ただしfは対数正規分布の確率密度関数であり、Fは累積分布関数である。  \n",
    "  また$t_u$は打ち切りなしのデータ、$t_{tc}$は打ち切りありのデータを指す。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 事前分布  \n",
    "  - こちらも原著にあるものを使用した。  \n",
    "  $\\mu$ ∼ Normal(6.5,25)   \n",
    "  $\\sigma^2$∼InverseGamma(6.5,23.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "stancode = \"\"\"\n",
    "data {\n",
    "  int N_u;\n",
    "  int N_tc;\n",
    "  vector[N_u] X_u;\n",
    "  vector[N_tc] X_tc;\n",
    "  int N_new;\n",
    "  vector[N_new] X_new;\n",
    "}\n",
    "\n",
    "parameters {\n",
    "  real mu;\n",
    "  real<lower=0> s_x;\n",
    "}\n",
    "\n",
    "model {\n",
    "   mu ~ normal(6.5, 25);\n",
    "   s_x ~ inv_gamma(6.5, 23.5);\n",
    "  for (n_u in 1:N_u)\n",
    "    X_u[n_u] ~ lognormal(mu, sqrt(s_x));\n",
    "  for (n_tc in 1:N_tc) {\n",
    "    target += lognormal_lccdf(X_tc[n_tc] | mu, sqrt(s_x));\n",
    "  }\n",
    "}\n",
    "\n",
    "generated quantities {\n",
    "  vector[N_new] lifetime;\n",
    "  for (n in 1:N_new){\n",
    "    lifetime[n] = exp(lognormal_lccdf(X_new[n] | mu, sqrt(s_x)));\n",
    "  }\n",
    "}\n",
    "\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# コンパイル済みのモデルがあれば読み込み。なければコンパイルし保存。\n",
    "try:\n",
    "    with open('./model_lifetime.pkl', \"rb\") as f:\n",
    "        stanmodel = pickle.load(f)\n",
    "except FileNotFoundError:\n",
    "    stanmodel = StanModel(model_code=stancode, model_name=\"model_lifetime\")\n",
    "    with open('./model_lifetime.pkl', \"wb\") as f:\n",
    "        pickle.dump(stanmodel, f)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. 推定"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.1. データの整形\n",
    "stan用にデータを変換する"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>value</th>\n",
       "      <th>trunc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>646</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1890</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1628</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>897</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>951</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    value  trunc\n",
       "1     646    NaN\n",
       "8    1890    NaN\n",
       "9    1628    NaN\n",
       "13    897    NaN\n",
       "28    951    NaN"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_u = df[df[\"trunc\"] != 1] # 打ち切りなしのデータ\n",
    "df_u.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>value</th>\n",
       "      <th>trunc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1085</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1795</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>727</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2238</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   value  trunc\n",
       "0   1085    1.0\n",
       "2   1795    1.0\n",
       "3    100    1.0\n",
       "4    727    1.0\n",
       "5   2238    1.0"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_tc = df[df[\"trunc\"] == 1] # 打ち切りありのデータ\n",
    "df_tc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "standata = {\"N_u\":df_u.shape[0], \"N_tc\":df_tc.shape[0], \"X_u\":df_u[\"value\"], \"X_tc\":df_tc[\"value\"], \n",
    "            \"N_new\":17, \"X_new\":np.linspace(0,3000,17)}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.2. MCMC実行"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "//anaconda/lib/python3.6/site-packages/pystan/misc.py:399: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  elif np.issubdtype(np.asarray(v).dtype, float):\n"
     ]
    }
   ],
   "source": [
    "fit = stanmodel.sampling(data = standata, iter=1000, warmup=400, seed=1234)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "MCMCにより得られた事後分布の概要を確認する"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>se_mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>2.5%</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>97.5%</th>\n",
       "      <th>n_eff</th>\n",
       "      <th>Rhat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mu</th>\n",
       "      <td>8.867294</td>\n",
       "      <td>0.014367</td>\n",
       "      <td>0.418379</td>\n",
       "      <td>8.170530</td>\n",
       "      <td>8.567788</td>\n",
       "      <td>8.818797</td>\n",
       "      <td>9.121048</td>\n",
       "      <td>9.794556</td>\n",
       "      <td>848.0</td>\n",
       "      <td>1.000724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>s_x</th>\n",
       "      <td>2.552790</td>\n",
       "      <td>0.030654</td>\n",
       "      <td>0.839495</td>\n",
       "      <td>1.393706</td>\n",
       "      <td>1.976292</td>\n",
       "      <td>2.382185</td>\n",
       "      <td>2.930969</td>\n",
       "      <td>4.738931</td>\n",
       "      <td>750.0</td>\n",
       "      <td>1.001302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[0]</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2400.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[1]</th>\n",
       "      <td>0.987904</td>\n",
       "      <td>0.000212</td>\n",
       "      <td>0.008531</td>\n",
       "      <td>0.965283</td>\n",
       "      <td>0.984513</td>\n",
       "      <td>0.990025</td>\n",
       "      <td>0.993808</td>\n",
       "      <td>0.997796</td>\n",
       "      <td>1618.0</td>\n",
       "      <td>1.000100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[2]</th>\n",
       "      <td>0.966793</td>\n",
       "      <td>0.000337</td>\n",
       "      <td>0.016511</td>\n",
       "      <td>0.925100</td>\n",
       "      <td>0.958056</td>\n",
       "      <td>0.969990</td>\n",
       "      <td>0.978743</td>\n",
       "      <td>0.990159</td>\n",
       "      <td>2400.0</td>\n",
       "      <td>0.999827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[3]</th>\n",
       "      <td>0.943423</td>\n",
       "      <td>0.000474</td>\n",
       "      <td>0.023225</td>\n",
       "      <td>0.887569</td>\n",
       "      <td>0.929435</td>\n",
       "      <td>0.947225</td>\n",
       "      <td>0.959962</td>\n",
       "      <td>0.978795</td>\n",
       "      <td>2400.0</td>\n",
       "      <td>0.999663</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[4]</th>\n",
       "      <td>0.919761</td>\n",
       "      <td>0.000594</td>\n",
       "      <td>0.029101</td>\n",
       "      <td>0.852901</td>\n",
       "      <td>0.901146</td>\n",
       "      <td>0.924013</td>\n",
       "      <td>0.940215</td>\n",
       "      <td>0.966329</td>\n",
       "      <td>2400.0</td>\n",
       "      <td>0.999595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[5]</th>\n",
       "      <td>0.896572</td>\n",
       "      <td>0.000702</td>\n",
       "      <td>0.034398</td>\n",
       "      <td>0.818538</td>\n",
       "      <td>0.874375</td>\n",
       "      <td>0.901355</td>\n",
       "      <td>0.920338</td>\n",
       "      <td>0.953318</td>\n",
       "      <td>2400.0</td>\n",
       "      <td>0.999591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[6]</th>\n",
       "      <td>0.874176</td>\n",
       "      <td>0.000801</td>\n",
       "      <td>0.039259</td>\n",
       "      <td>0.786193</td>\n",
       "      <td>0.849264</td>\n",
       "      <td>0.878338</td>\n",
       "      <td>0.901993</td>\n",
       "      <td>0.940028</td>\n",
       "      <td>2400.0</td>\n",
       "      <td>0.999626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[7]</th>\n",
       "      <td>0.852699</td>\n",
       "      <td>0.000959</td>\n",
       "      <td>0.043762</td>\n",
       "      <td>0.756956</td>\n",
       "      <td>0.824679</td>\n",
       "      <td>0.856178</td>\n",
       "      <td>0.883870</td>\n",
       "      <td>0.927768</td>\n",
       "      <td>2081.0</td>\n",
       "      <td>0.999682</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[8]</th>\n",
       "      <td>0.832173</td>\n",
       "      <td>0.001090</td>\n",
       "      <td>0.047954</td>\n",
       "      <td>0.728689</td>\n",
       "      <td>0.801676</td>\n",
       "      <td>0.835574</td>\n",
       "      <td>0.866551</td>\n",
       "      <td>0.915562</td>\n",
       "      <td>1935.0</td>\n",
       "      <td>0.999750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[9]</th>\n",
       "      <td>0.812588</td>\n",
       "      <td>0.001264</td>\n",
       "      <td>0.051868</td>\n",
       "      <td>0.700541</td>\n",
       "      <td>0.780121</td>\n",
       "      <td>0.815259</td>\n",
       "      <td>0.849903</td>\n",
       "      <td>0.903410</td>\n",
       "      <td>1684.0</td>\n",
       "      <td>0.999822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[10]</th>\n",
       "      <td>0.793908</td>\n",
       "      <td>0.001405</td>\n",
       "      <td>0.055528</td>\n",
       "      <td>0.675550</td>\n",
       "      <td>0.760103</td>\n",
       "      <td>0.796174</td>\n",
       "      <td>0.833977</td>\n",
       "      <td>0.892340</td>\n",
       "      <td>1561.0</td>\n",
       "      <td>0.999895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[11]</th>\n",
       "      <td>0.776091</td>\n",
       "      <td>0.001542</td>\n",
       "      <td>0.058953</td>\n",
       "      <td>0.650647</td>\n",
       "      <td>0.739242</td>\n",
       "      <td>0.778386</td>\n",
       "      <td>0.818558</td>\n",
       "      <td>0.881099</td>\n",
       "      <td>1461.0</td>\n",
       "      <td>0.999966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[12]</th>\n",
       "      <td>0.759086</td>\n",
       "      <td>0.001674</td>\n",
       "      <td>0.062160</td>\n",
       "      <td>0.627497</td>\n",
       "      <td>0.719941</td>\n",
       "      <td>0.760938</td>\n",
       "      <td>0.803771</td>\n",
       "      <td>0.870099</td>\n",
       "      <td>1378.0</td>\n",
       "      <td>1.000034</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[13]</th>\n",
       "      <td>0.742847</td>\n",
       "      <td>0.001800</td>\n",
       "      <td>0.065163</td>\n",
       "      <td>0.606170</td>\n",
       "      <td>0.700958</td>\n",
       "      <td>0.743998</td>\n",
       "      <td>0.789632</td>\n",
       "      <td>0.859942</td>\n",
       "      <td>1310.0</td>\n",
       "      <td>1.000099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[14]</th>\n",
       "      <td>0.727325</td>\n",
       "      <td>0.001920</td>\n",
       "      <td>0.067978</td>\n",
       "      <td>0.586425</td>\n",
       "      <td>0.683444</td>\n",
       "      <td>0.728215</td>\n",
       "      <td>0.775899</td>\n",
       "      <td>0.850184</td>\n",
       "      <td>1254.0</td>\n",
       "      <td>1.000160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[15]</th>\n",
       "      <td>0.712476</td>\n",
       "      <td>0.002034</td>\n",
       "      <td>0.070615</td>\n",
       "      <td>0.566867</td>\n",
       "      <td>0.666964</td>\n",
       "      <td>0.713373</td>\n",
       "      <td>0.763081</td>\n",
       "      <td>0.840086</td>\n",
       "      <td>1205.0</td>\n",
       "      <td>1.000218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lifetime[16]</th>\n",
       "      <td>0.698257</td>\n",
       "      <td>0.002142</td>\n",
       "      <td>0.073087</td>\n",
       "      <td>0.547913</td>\n",
       "      <td>0.650799</td>\n",
       "      <td>0.699134</td>\n",
       "      <td>0.750325</td>\n",
       "      <td>0.830725</td>\n",
       "      <td>1164.0</td>\n",
       "      <td>1.000272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lp__</th>\n",
       "      <td>-34.867070</td>\n",
       "      <td>0.032991</td>\n",
       "      <td>1.013640</td>\n",
       "      <td>-37.616381</td>\n",
       "      <td>-35.282798</td>\n",
       "      <td>-34.573338</td>\n",
       "      <td>-34.122728</td>\n",
       "      <td>-33.866237</td>\n",
       "      <td>944.0</td>\n",
       "      <td>1.002463</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   mean   se_mean        sd       2.5%        25%        50%  \\\n",
       "mu             8.867294  0.014367  0.418379   8.170530   8.567788   8.818797   \n",
       "s_x            2.552790  0.030654  0.839495   1.393706   1.976292   2.382185   \n",
       "lifetime[0]    1.000000  0.000000  0.000000   1.000000   1.000000   1.000000   \n",
       "lifetime[1]    0.987904  0.000212  0.008531   0.965283   0.984513   0.990025   \n",
       "lifetime[2]    0.966793  0.000337  0.016511   0.925100   0.958056   0.969990   \n",
       "lifetime[3]    0.943423  0.000474  0.023225   0.887569   0.929435   0.947225   \n",
       "lifetime[4]    0.919761  0.000594  0.029101   0.852901   0.901146   0.924013   \n",
       "lifetime[5]    0.896572  0.000702  0.034398   0.818538   0.874375   0.901355   \n",
       "lifetime[6]    0.874176  0.000801  0.039259   0.786193   0.849264   0.878338   \n",
       "lifetime[7]    0.852699  0.000959  0.043762   0.756956   0.824679   0.856178   \n",
       "lifetime[8]    0.832173  0.001090  0.047954   0.728689   0.801676   0.835574   \n",
       "lifetime[9]    0.812588  0.001264  0.051868   0.700541   0.780121   0.815259   \n",
       "lifetime[10]   0.793908  0.001405  0.055528   0.675550   0.760103   0.796174   \n",
       "lifetime[11]   0.776091  0.001542  0.058953   0.650647   0.739242   0.778386   \n",
       "lifetime[12]   0.759086  0.001674  0.062160   0.627497   0.719941   0.760938   \n",
       "lifetime[13]   0.742847  0.001800  0.065163   0.606170   0.700958   0.743998   \n",
       "lifetime[14]   0.727325  0.001920  0.067978   0.586425   0.683444   0.728215   \n",
       "lifetime[15]   0.712476  0.002034  0.070615   0.566867   0.666964   0.713373   \n",
       "lifetime[16]   0.698257  0.002142  0.073087   0.547913   0.650799   0.699134   \n",
       "lp__         -34.867070  0.032991  1.013640 -37.616381 -35.282798 -34.573338   \n",
       "\n",
       "                    75%      97.5%   n_eff      Rhat  \n",
       "mu             9.121048   9.794556   848.0  1.000724  \n",
       "s_x            2.930969   4.738931   750.0  1.001302  \n",
       "lifetime[0]    1.000000   1.000000  2400.0       NaN  \n",
       "lifetime[1]    0.993808   0.997796  1618.0  1.000100  \n",
       "lifetime[2]    0.978743   0.990159  2400.0  0.999827  \n",
       "lifetime[3]    0.959962   0.978795  2400.0  0.999663  \n",
       "lifetime[4]    0.940215   0.966329  2400.0  0.999595  \n",
       "lifetime[5]    0.920338   0.953318  2400.0  0.999591  \n",
       "lifetime[6]    0.901993   0.940028  2400.0  0.999626  \n",
       "lifetime[7]    0.883870   0.927768  2081.0  0.999682  \n",
       "lifetime[8]    0.866551   0.915562  1935.0  0.999750  \n",
       "lifetime[9]    0.849903   0.903410  1684.0  0.999822  \n",
       "lifetime[10]   0.833977   0.892340  1561.0  0.999895  \n",
       "lifetime[11]   0.818558   0.881099  1461.0  0.999966  \n",
       "lifetime[12]   0.803771   0.870099  1378.0  1.000034  \n",
       "lifetime[13]   0.789632   0.859942  1310.0  1.000099  \n",
       "lifetime[14]   0.775899   0.850184  1254.0  1.000160  \n",
       "lifetime[15]   0.763081   0.840086  1205.0  1.000218  \n",
       "lifetime[16]   0.750325   0.830725  1164.0  1.000272  \n",
       "lp__         -34.122728 -33.866237   944.0  1.002463  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit.summary()をDataFrameに整形する\n",
    "fit_summary = pd.DataFrame(data = fit.summary()[\"summary\"], index = fit.summary()[\"summary_rownames\"], \n",
    "                          columns= fit.summary()[\"summary_colnames\"])\n",
    "fit_summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 事後分布およびトレースプロットの確認\n",
    "sns.set(rc={\"figure.figsize\":(12,5)}) # サイズの調整に必要\n",
    "plt.style.use('ggplot')\n",
    "fit.plot(pars=[\"mu\", \"s_x\"])\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. 結果\n",
    "生存関数(survival function)の90%信用区間を示す。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "ms = fit.extract()\n",
    "lifeDF = pd.DataFrame(data=ms[\"lifetime\"], columns=standata[\"X_new\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = lifeDF.mean(axis=0).plot(marker = \"o\")\n",
    "ax = lifeDF.quantile(0.05, axis=0).plot(color=\"k\", linestyle=\"--\", marker = \".\", alpha=0.3)\n",
    "ax = lifeDF.quantile(0.95, axis=0).plot(color=\"k\", linestyle=\"--\", marker = \".\", alpha=0.3)\n",
    "ax.fill_between(standata[\"X_new\"], lifeDF.quantile(0.05, axis=0), lifeDF.quantile(0.95, axis=0), \n",
    "                alpha=0.3, color=\"c1\")#, interpolate=True)\n",
    "\n",
    "ax.set_xlabel('lifetime[h]')\n",
    "ax.set_ylabel('survival function')\n",
    "ax.set_title('90% credible interval of survivial function 1-F(T)')\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- この結果から、例えば90%の製品が500時間は稼働すると推定することができる。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 参考文献\n",
    "[1].  Hamada, M. S., Wilson, A., Reese, C. S., & Martz, H. (2008). Bayesian reliability. Springer Science & Business Media."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda env:anaconda]",
   "language": "python",
   "name": "conda-env-anaconda-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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}