reouno/small-edge-search-01-use-other-indexes.ipynb

## small-edge-search-01-use-other-indexes.ipynb
{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 小さいエッジの大量探索第一段階\n",
    "\n",
    "常にターゲットはE-mini S&P500先物（**ES**）。足の単位は自由。予測の方法も自由。予測の内容も自由（次の足の上昇を予測するとか、N足以内の下落を予測するとか）。\n",
    "\n",
    "ただし、基本的に**点推定ではなく分布推定**をする。点推定の場合、単に上昇確率=◯◯%などとなるがこの場合、1ポイント以上上昇とか下落が10ポイント以内におさまるとかの確率を知りたいときにいちいち元のデータから計算し直す必要がある。これに対して分布を推定しておけば、後はその分布を元にして確率を計算できるため実際に使用するときに知りたい確率を容易に計算できて便利。\n",
    "\n",
    "\n",
    "- 上昇した足の翌足の収益率分布\n",
    "- 連続でN足下落した翌足の収益率分布\n",
    "- 銘柄Aが上昇した足の翌足でのESの収益率分布\n",
    "- 銘柄Aが下落した足の翌足でのESの収益率分布\n",
    "- 銘柄Aが2日連続下落した足の翌足でのESの収益率分布\n",
    "\n",
    "などの分布を片っぱしから推定していく。このルールだけを与えたら自動で分布を推定して結果を記録していくような仕組みがあると良い。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 手始めに上昇した日の翌日の収益率分布を全体の分布と比較する\n",
    "\n",
    "仮説：上昇した日の翌日はその反動で下落しやすいと仮定すると、上昇した日の翌日の収益率分布は全期間の収益率分布よりも左にずれると予想できる。\n",
    "\n",
    "## 結論\n",
    "\n",
    "結果については一番下のセルを参照。\n",
    "\n",
    "上昇日の翌日の収益率は全期間の収益率とほぼ同じであり、単純に上昇日の翌日に売るとか買うというルールでランダム売買戦略より大きい収益を上げることはできないだろう。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 準備"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'en_US.UTF-8'"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pandas.api.types import CategoricalDtype\n",
    "import matplotlib as mpl\n",
    "mpl.rcParams['font.family'] = 'sans-serif'\n",
    "mpl.rcParams['font.sans-serif'] = ['Hiragino Maru Gothic Pro', 'Yu Gothic', 'Meirio', 'Takao', 'IPAexGothic', 'IPAPGothic', 'VL PGothic', 'Noto Sans CJK JP']\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from scipy import stats\n",
    "import datetime as dt\n",
    "from dateutil.relativedelta import relativedelta\n",
    "import locale\n",
    "\n",
    "# 月や曜日を英語で取得するためこの設定をしておく\n",
    "locale.setlocale(locale.LC_TIME, 'en_US.UTF-8')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### E-Mini S&P500先物と金先物データ読み込み\n",
    "\n",
    "データは、[TradeStationのDesktop Platform](https://www.tradestation.com/platforms-and-tools/desktop/)を使って出力したCSVファイル。公開は禁止されているので公開できないが、Quandlからも[E-mini S&P500先物](https://www.quandl.com/data/CHRIS/CME_ES1-E-mini-S-P-500-Futures-Continuous-Contract-1-ES1-Front-Month)の最新の日足データ（今回の分析で使っているのと同じ）をAPI経由で取得できるので、そのデータを使用可能。ただしquandl経由で取得したDFは若干構造が違うから、以下のセルは多少修正が必要。\n",
    "- quandl経由で取得したDFはすでにDatetimeindexになっている。また、終値は'Close'ではなく'Last'。おそらくこの2点だけ違う。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/leo/src/pyproject/py-envs/py383env/lib/python3.6/site-packages/pandas/core/series.py:679: RuntimeWarning: divide by zero encountered in log\n",
      "  result = getattr(ufunc, method)(*inputs, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "df_tmp = pd.read_csv('data/e-mini-sp500-200530/e-mini-sp500-daily.csv')\n",
    "\n",
    "# datetime indexに変換\n",
    "def to_datetime_index(df):\n",
    "    # DateTime列を追加\n",
    "    df['datetime'] = (df['Date'] + '-' + df['Time']).map(lambda s: dt.datetime.strptime(s, '%m/%d/%Y-%H:%M'))\n",
    "    df = df.set_index('datetime', drop=True)\n",
    "    df = df.drop(columns=['Date', 'Time'])\n",
    "    return df\n",
    "\n",
    "df = to_datetime_index(df_tmp)\n",
    "\n",
    "# 対数変換した列を追加\n",
    "def add_log_values(df):\n",
    "    df['logO'] = np.log(df['Open'])\n",
    "    df['logH'] = np.log(df['High'])\n",
    "    df['logL'] = np.log(df['Low'])\n",
    "    df['logC'] = np.log(df['Close'])\n",
    "    df['logV'] = np.log(df['Vol'])\n",
    "    df['logOI'] = np.log(df['OI'])\n",
    "\n",
    "_ = add_log_values(df)"
   ]
  },
  {
   "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>Date</th>\n",
       "      <th>Time</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Vol</th>\n",
       "      <th>OI</th>\n",
       "      <th>datetime</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>09/11/1997</td>\n",
       "      <td>17:00</td>\n",
       "      <td>1071.25</td>\n",
       "      <td>1082.25</td>\n",
       "      <td>1062.75</td>\n",
       "      <td>1068.50</td>\n",
       "      <td>11825</td>\n",
       "      <td>2909</td>\n",
       "      <td>1997-09-11 17:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>09/12/1997</td>\n",
       "      <td>17:00</td>\n",
       "      <td>1070.50</td>\n",
       "      <td>1089.00</td>\n",
       "      <td>1066.00</td>\n",
       "      <td>1071.25</td>\n",
       "      <td>9759</td>\n",
       "      <td>4059</td>\n",
       "      <td>1997-09-12 17:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Date   Time     Open     High      Low    Close    Vol    OI  \\\n",
       "0  09/11/1997  17:00  1071.25  1082.25  1062.75  1068.50  11825  2909   \n",
       "1  09/12/1997  17:00  1070.50  1089.00  1066.00  1071.25   9759  4059   \n",
       "\n",
       "             datetime  \n",
       "0 1997-09-11 17:00:00  \n",
       "1 1997-09-12 17:00:00  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 参考：生のデータフレーム\n",
    "df_tmp.head(2)"
   ]
  },
  {
   "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>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Vol</th>\n",
       "      <th>OI</th>\n",
       "      <th>logO</th>\n",
       "      <th>logH</th>\n",
       "      <th>logL</th>\n",
       "      <th>logC</th>\n",
       "      <th>logV</th>\n",
       "      <th>logOI</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datetime</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1997-09-11 17:00:00</th>\n",
       "      <td>1071.25</td>\n",
       "      <td>1082.25</td>\n",
       "      <td>1062.75</td>\n",
       "      <td>1068.50</td>\n",
       "      <td>11825</td>\n",
       "      <td>2909</td>\n",
       "      <td>6.976581</td>\n",
       "      <td>6.986797</td>\n",
       "      <td>6.968615</td>\n",
       "      <td>6.974011</td>\n",
       "      <td>9.377971</td>\n",
       "      <td>7.975565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1997-09-12 17:00:00</th>\n",
       "      <td>1070.50</td>\n",
       "      <td>1089.00</td>\n",
       "      <td>1066.00</td>\n",
       "      <td>1071.25</td>\n",
       "      <td>9759</td>\n",
       "      <td>4059</td>\n",
       "      <td>6.975881</td>\n",
       "      <td>6.993015</td>\n",
       "      <td>6.971669</td>\n",
       "      <td>6.976581</td>\n",
       "      <td>9.185945</td>\n",
       "      <td>8.308692</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close    Vol    OI  \\\n",
       "datetime                                                               \n",
       "1997-09-11 17:00:00  1071.25  1082.25  1062.75  1068.50  11825  2909   \n",
       "1997-09-12 17:00:00  1070.50  1089.00  1066.00  1071.25   9759  4059   \n",
       "\n",
       "                         logO      logH      logL      logC      logV  \\\n",
       "datetime                                                                \n",
       "1997-09-11 17:00:00  6.976581  6.986797  6.968615  6.974011  9.377971   \n",
       "1997-09-12 17:00:00  6.975881  6.993015  6.971669  6.976581  9.185945   \n",
       "\n",
       "                        logOI  \n",
       "datetime                       \n",
       "1997-09-11 17:00:00  7.975565  \n",
       "1997-09-12 17:00:00  8.308692  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 参考：対数変換データ追加後のデータフレーム\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 価格、対数価格、価格階差、対数差収益率（100倍）のDFを作成"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/leo/src/pyproject/py-envs/py383env/lib/python3.6/site-packages/ipykernel_launcher.py:6: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "def to_log_return_ratio_df(df):\n",
    "    diff_df = df.diff()\n",
    "    close_df = df[['Close', 'logC']]\n",
    "    diff_df = diff_df.rename(columns={'Close': 'CloseDiff', 'logC': 'logCDiff'})\n",
    "    close_diff_df = diff_df[['CloseDiff', 'logCDiff']]\n",
    "    close_diff_df['logCDiff'] = close_diff_df['logCDiff'] * 100\n",
    "    rr_df = pd.concat([close_df, close_diff_df], axis=1)\n",
    "    rr_df = rr_df.dropna()\n",
    "    return rr_df\n",
    "\n",
    "rr_df = to_log_return_ratio_df(df)"
   ]
  },
  {
   "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>Close</th>\n",
       "      <th>logC</th>\n",
       "      <th>CloseDiff</th>\n",
       "      <th>logCDiff</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datetime</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1997-09-12 17:00:00</th>\n",
       "      <td>1071.25</td>\n",
       "      <td>6.976581</td>\n",
       "      <td>2.75</td>\n",
       "      <td>0.257040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1997-09-15 17:00:00</th>\n",
       "      <td>1083.75</td>\n",
       "      <td>6.988183</td>\n",
       "      <td>12.50</td>\n",
       "      <td>1.160106</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Close      logC  CloseDiff  logCDiff\n",
       "datetime                                                   \n",
       "1997-09-12 17:00:00  1071.25  6.976581       2.75  0.257040\n",
       "1997-09-15 17:00:00  1083.75  6.988183      12.50  1.160106"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 参考：対数差分をとった後は、初日のデータはなくなる\n",
    "rr_df.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 分布の推定と比較"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 「翌足の対数差収益率」という列を追加\n",
    "rr_df2d = rr_df.rename(columns={'logCDiff': 'logCDiff0'})\n",
    "# datetimeindexなので、一旦インデックスを消さないとうまく結合できない\n",
    "# しかしto_numpy()は遅いから別の方法でやりたい\n",
    "rr_df2d['logCDiff1'] = rr_df2d['logCDiff0'][1:].append(pd.Series([np.nan]*1)).to_numpy()\n",
    "\n",
    "# logCDiff1列の最後はNaNになるためその行を除外（最後の足の翌足のデータは無いから）\n",
    "rr_df2d = rr_df2d.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>Close</th>\n",
       "      <th>logC</th>\n",
       "      <th>CloseDiff</th>\n",
       "      <th>logCDiff0</th>\n",
       "      <th>logCDiff1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datetime</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1997-09-12 17:00:00</th>\n",
       "      <td>1071.25</td>\n",
       "      <td>6.976581</td>\n",
       "      <td>2.75</td>\n",
       "      <td>0.257040</td>\n",
       "      <td>1.160106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1997-09-15 17:00:00</th>\n",
       "      <td>1083.75</td>\n",
       "      <td>6.988183</td>\n",
       "      <td>12.50</td>\n",
       "      <td>1.160106</td>\n",
       "      <td>2.258050</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Close      logC  CloseDiff  logCDiff0  logCDiff1\n",
       "datetime                                                               \n",
       "1997-09-12 17:00:00  1071.25  6.976581       2.75   0.257040   1.160106\n",
       "1997-09-15 17:00:00  1083.75  6.988183      12.50   1.160106   2.258050"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 参考：logCDiff0は当足の収益率、logCDiff1は翌足の収益率\n",
    "# 当足の収益率がプラスだった場合の翌足の収益率分布が知りたい\n",
    "rr_df2d.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "上昇した日の割合 = 3045 / 5727 = 53.17%\n"
     ]
    }
   ],
   "source": [
    "up_df = rr_df2d[rr_df2d['logCDiff0'] > 0]\n",
    "n_all = rr_df2d.shape[0]\n",
    "n_up = up_df.shape[0]\n",
    "print(f'上昇した日の割合 = {n_up} / {n_all} = {n_up / n_all * 100:.02f}%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "全期間の対数差収益率 のt分布パラメータ\n",
      "df=2.3669323230150696, loc=0.07086735101302821, scale=0.7277745201048187\n",
      "上昇日の翌日の収益率 のt分布パラメータ\n",
      "df=2.360011616962992, loc=0.04749138631505906, scale=0.6510313403914763\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1296 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(3, 1, figsize=(20, 18))\n",
    "\n",
    "# 全期間の収益率分布を描く\n",
    "# 上昇した日の翌日の収益率分布を描く\n",
    "# 両者を重ねる\n",
    "# 各々をt分布と仮定してパラメータ推定（この時、νは同じ値にしておいた方が良いのか？？）\n",
    "plot_data = [rr_df2d['logCDiff0'], up_df['logCDiff1']]\n",
    "titles = ['全期間の対数差収益率', '上昇日の翌日の収益率']\n",
    "colors=['tab:green','tab:pink']\n",
    "\n",
    "# x軸の範囲を広い方（全期間）に合わせる\n",
    "xmin = plot_data[0].min()\n",
    "xmax = plot_data[0].max()\n",
    "\n",
    "# t分布の描画範囲\n",
    "xs = np.linspace(xmin, xmax, 300)\n",
    "\n",
    "t_params = []\n",
    "t_ys = []\n",
    "for i in range(2):\n",
    "    sns.histplot(plot_data[i], kde=False, stat='density', color='lightblue', ax=ax[i])\n",
    "\n",
    "    # t分布の当てはめ\n",
    "    t_params.append(stats.t.fit(plot_data[i]))\n",
    "    t_ys.append(stats.t.pdf(xs, df=t_params[i][0], loc=t_params[i][1], scale=t_params[i][2]))\n",
    "\n",
    "    ax[i].set_xlim(xmin, xmax)\n",
    "    ax[i].plot(xs, t_ys[i], color=colors[i])\n",
    "\n",
    "    ax[i].set_title(titles[i], fontweight='semibold', fontsize=16)\n",
    "\n",
    "# 推定分布を重ねて比較\n",
    "ax[2].plot(xs, t_ys[0], label=titles[0], color=colors[0])\n",
    "ax[2].plot(xs, t_ys[1], label=titles[1], color=colors[1])\n",
    "ax[2].set_xlim(xmin, xmax)\n",
    "ax[2].legend()\n",
    "\n",
    "# パラメータ推定値の確認\n",
    "for i in range(2):\n",
    "    print(titles[i], 'のt分布パラメータ')\n",
    "    print(f'df={t_params[i][0]}, loc={t_params[i][1]}, scale={t_params[i][2]}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 結果"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "全期間の収益率分布と上昇日の翌日の収益率分布の間に差はほとんど見られなかった。分布の重なり具合から見て、統計検定の必要はないだろう。\n",
    "\n",
    "\n",
    "#### 補足：t分布のパラメータνについて\n",
    "\n",
    "t分布を比較する際、パラメータνを固定した方が良いのか。当てはまりの良さを比較する場合は、対数尤度やAICやBICなどの指標を使うから、パラメータを固定する必要はないし、正規分布とt分布の比較など他分布同士でも良い。また、異なる分布であってもそれを元に計算された「◯◯以上上昇する確率」などの具体的な条件付き確率についても比較できる。\n",
    "\n",
    "しかし、平均や分散を比較する場合、例えば正規分布とt分布の平均や分散を比較することはできない。なぜならこの2つはそもそも異なる分布であり、同じデータに対してそれぞれの分布を当てはめた場合でも平均と分散はそれぞれ異なる値になるから。\n",
    "\n",
    "とすると、νの値が異なる2つのt分布がある時、平均や分散を直接比較することはできない。従って、以下のようになるだろう。\n",
    "\n",
    "- 当てはまりの良さを比較する場合：νを固定する必要はない（異なる分布同士を比較可能）\n",
    "- 各々の分布から計算された確率の比較：νを固定する必要はない（そもそも分布の比較ではない）\n",
    "- 平均と分散の値を直接比較する場合：νを固定する必要がある"
   ]
  }
 ],
 "metadata": {
  "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}