2_Model Evaluation (Q1).ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "4627b24b-0946-4837-8ada-d8b12f243dac",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "pd.set_option('display.max_row', 150)\n",
    "import warnings\n",
    "from pandas.core.common import SettingWithCopyWarning\n",
    "warnings.simplefilter(action=\"ignore\", category=SettingWithCopyWarning)\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "b00d7e2a-e00e-4ef8-81fb-be8e972d74e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "result = pd.read_csv(\"./eval_result/predict_result.csv\")\n",
    "result = result[result[\"model\"]!=\"svm\"]\n",
    "result = result[result[\"model\"]!=\"et\"]\n",
    "\n",
    "result2 = pd.read_csv(\"./eval_result/predict_result2.csv\")\n",
    "result2 = result2[result2[\"model\"]!=\"svm\"]\n",
    "result2 = result2[result2[\"model\"]!=\"et\"]\n",
    "\n",
    "result3 = pd.read_csv(\"./eval_result/predict_result3.csv\")\n",
    "result3 = result3[result3[\"model\"]!=\"svm\"]\n",
    "result3 = result3[result3[\"model\"]!=\"et\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "f46d4c95-f914-4062-9e2b-c5ddf91094fd",
   "metadata": {},
   "outputs": [],
   "source": [
    "result2 = pd.concat([result2, result3]).reset_index(drop = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "9ed599cd-0d61-48ef-a1e3-f7bc9776b581",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['GC', 'BZ', 'CL', 'VX', 'MGC', 'UB', 'MME'], dtype=object)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result3[\"data\"].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e5989000-3be9-4a6c-9d96-3ea12e62c0cd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "9cd1ceb9-f6fc-4f96-a4a4-4b2dcc29e71c",
   "metadata": {},
   "source": [
    "### 3.2.1.1. Model Evaluation (Q1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "cc38c18d-8dd4-42e5-92dc-2bf276d97c35",
   "metadata": {},
   "outputs": [],
   "source": [
    "def model_eval(path, coin_name, model_name, time):\n",
    "    \n",
    "    \"\"\"\n",
    "    path : risk result path\n",
    "    coin_name : name of coin \n",
    "    model_name : name of model \n",
    "    time : e.g., 4h, 8h, 12h, 1d\n",
    "    \"\"\"\n",
    "    \n",
    "\n",
    "    time_norm = pd.read_csv(f\"./eval_result/{path}.csv\")\n",
    "\n",
    "    time_norm = time_norm[time_norm[\"data\"] == coin_name]\n",
    "    time_norm = time_norm[time_norm[\"model\"] ==  model_name]\n",
    "\n",
    "\n",
    "    mask_targer_label = time_norm[\"label\"].isin([\"Target_Label\"])\n",
    "    time_norm_target = time_norm[mask_targer_label]\n",
    "    time_norm = time_norm[~mask_targer_label]\n",
    "\n",
    "\n",
    "    post_label = [ 'Target_2.0post_Label',\n",
    "           'Target_2.2post_Label', 'Target_2.4post_Label',\n",
    "           'Target_2.6post_Label', 'Target_2.8post_Label',\n",
    "           'Target_3.0post_Label', 'Target_3.2post_Label',\n",
    "           'Target_3.4post_Label', 'Target_3.6post_Label',\n",
    "           'Target_3.8post_Label', 'Target_4.0post_Label']\n",
    "\n",
    "    mask2 = time_norm[\"label\"].isin(post_label)\n",
    "\n",
    "    post_data = time_norm[mask2]\n",
    "    new_data = time_norm[~mask2]\n",
    "\n",
    "    time_norm_target.iloc[:, 4:9] = time_norm_target.iloc[:, 4:9].astype(float).copy() \n",
    "    new_data.iloc[:, 4:9] = new_data.iloc[:, 4:9].astype(float).round(4).copy()\n",
    "\n",
    "    time_norm_target = time_norm_target.iloc[:, :9].reset_index().iloc[:, 1:]\n",
    "    new_data = new_data.iloc[:, :9].reset_index().iloc[:, 1:]\n",
    "    new_data = new_data[new_data[\"time\"] == time].reset_index().iloc[:, 1:]\n",
    "    new_data = new_data.sort_values(\"label\").reset_index().iloc[:, 1:]\n",
    "\n",
    "    time_norm_target_value = pd.DataFrame([time_norm_target[time_norm_target[\"time\"] == time].iloc[:, 4:].reset_index().iloc[:, 1:].values[0] for i in range(new_data[new_data[\"time\"] == time].shape[0])])\n",
    "    time_norm_target_value.columns = [\"acc\",\"auc\", \"mcc\", \"precision\", \"recall\"]\n",
    "    \n",
    "    percent = (((new_data.iloc[:, 4:] / time_norm_target_value)-1 ) * 100).round(4)\n",
    "    percent.columns = [\"acc_p\",\"auc_p\", \"mcc_p\", \"precision_p\", \"recall_p\"]\n",
    "    \n",
    "    time_norm_target_value = pd.concat([new_data.iloc[:, :4], time_norm_target_value], axis = 1)\n",
    "    \n",
    "    output =  pd.concat([time_norm_target_value,new_data ]).reset_index().iloc[:, 1:]\n",
    "    output = output.drop([\"data\",\"time\",\"model\"], axis = 1)\n",
    "    \n",
    "    melt_data = time_norm_target_value.iloc[0:1, :][[\"label\",\"acc\",\"auc\"]].copy()\n",
    "    melt_data = melt_data.melt(\"label\")\n",
    "    melt_data[\"label\"] = \"Target_Label\"\n",
    "    \n",
    "    new_data = new_data[[\"label\",\"acc\",\"auc\"]]\n",
    "    new_data = new_data.melt([\"label\"])\n",
    "    \n",
    "    \n",
    "    \n",
    "    return new_data, melt_data #pd.concat([new_data, melt_data]).reset_index().iloc[:, 1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "6e6da415-f316-4231-83e7-3f52ffd454bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "cat_4h , melt_4h     =model_eval(\"predict_result\", \"BTC\", \"cat\", \"4h\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "fd161f7d-3487-4972-a82d-503d32ea1703",
   "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>label</th>\n",
       "      <th>variable</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Target_Label</td>\n",
       "      <td>acc</td>\n",
       "      <td>0.546713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Target_Label</td>\n",
       "      <td>auc</td>\n",
       "      <td>0.546742</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          label variable     value\n",
       "0  Target_Label      acc  0.546713\n",
       "1  Target_Label      auc  0.546742"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "melt_4h # ST_\\delta - NT"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65f77fb0-f8fb-4240-8933-c2a44d9ad337",
   "metadata": {},
   "source": [
    "## Q1 plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "2f05ffe9-f3df-4496-801a-f459d64a53ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "def all_models(predict_result, coin):\n",
    "    cat_4h , melt_4h     =model_eval(predict_result, coin, \"cat\", \"4h\")\n",
    "    rf_4h , melt_rf_4h   =model_eval(predict_result, coin, \"rf\", \"4h\")\n",
    "    mlp_4h , melt_mlp_4h =model_eval(predict_result, coin, \"mlp\", \"4h\")\n",
    "    knn_4h , melt_knn_4h =model_eval(predict_result, coin, \"knn\", \"4h\")\n",
    "    et_4h , melt_et_4h   =model_eval(predict_result, coin, \"et\", \"4h\")\n",
    "    xg_4h , melt_xg_4h   =model_eval(predict_result, coin, \"xg\", \"4h\")\n",
    "\n",
    "\n",
    "    auc_cat_4h_data = cat_4h[cat_4h[\"variable\"] == \"auc\"][\"value\"].values - melt_4h[melt_4h[\"variable\"] == \"auc\"][\"value\"].values\n",
    "    auc_rf_4h_data = rf_4h[rf_4h[\"variable\"] == \"auc\"][\"value\"].values - melt_rf_4h[melt_rf_4h[\"variable\"] == \"auc\"][\"value\"].values\n",
    "    auc_mlp_4h_data = mlp_4h[mlp_4h[\"variable\"] == \"auc\"][\"value\"].values - melt_mlp_4h[melt_mlp_4h[\"variable\"] == \"auc\"][\"value\"].values\n",
    "    auc_knn_4h_data = knn_4h[knn_4h[\"variable\"] == \"auc\"][\"value\"].values - melt_knn_4h[melt_knn_4h[\"variable\"] == \"auc\"][\"value\"].values\n",
    "    auc_et_4h_data = et_4h[et_4h[\"variable\"] == \"auc\"][\"value\"].values - melt_et_4h[melt_et_4h[\"variable\"] == \"auc\"][\"value\"].values\n",
    "    auc_xg_4h_data = xg_4h[xg_4h[\"variable\"] == \"auc\"][\"value\"].values - melt_xg_4h[melt_xg_4h[\"variable\"] == \"auc\"][\"value\"].values\n",
    "\n",
    "    auc_data_4h  = pd.DataFrame([auc_cat_4h_data,  auc_rf_4h_data,  auc_mlp_4h_data, auc_knn_4h_data, auc_et_4h_data, auc_xg_4h_data]).melt().iloc[:, 1]\n",
    "    \n",
    "    auc_data =  pd.concat([auc_data_4h], axis = 1)\n",
    "    auc_data.columns = [\"4h\"]\n",
    "    auc_data[\"delta\"] = ['d2.0','d2.2', 'd2.4', 'd2.6', 'd2.8', 'd3.0', 'd3.2', 'd3.4', 'd3.6', 'd3.8', 'd4.0'] * 6\n",
    "    \n",
    "    auc_data = auc_data.melt(\"delta\")\n",
    "\n",
    "    return auc_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "b6096423-3217-42ef-8287-d48d796a0d6c",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "btc_auc = all_models(\"predict_result\",\"BTC\")\n",
    "eth_auc = all_models(\"predict_result\",\"ETH\")\n",
    "doge_auc= all_models(\"predict_result\",\"DOGE\")\n",
    "sol_auc = all_models(\"predict_result\",\"SOL\")\n",
    "bnb_auc = all_models(\"predict_result\",\"BNB\")\n",
    "ada_auc = all_models(\"predict_result\",\"ADA\")\n",
    "xrp_auc = all_models(\"predict_result\",\"XRP\")\n",
    "\n",
    "total_auc = pd.concat([btc_auc, eth_auc, doge_auc, sol_auc, bnb_auc, ada_auc, xrp_auc]).reset_index().iloc[:,1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "a971772a-75d8-4a8e-a392-94bd11a82aeb",
   "metadata": {},
   "outputs": [],
   "source": [
    "YM_auc = all_models(\"predict_result2\",\"YM\")\n",
    "NQ_auc = all_models(\"predict_result2\",\"NQ\")\n",
    "ES_auc = all_models(\"predict_result2\",\"ES\")\n",
    "PA_auc = all_models(\"predict_result2\",\"PA\")\n",
    "EW_auc = all_models(\"predict_result2\",\"EW\")\n",
    "UB_auc = all_models(\"predict_result3\",\"UB\")\n",
    "CL_auc = all_models(\"predict_result3\",\"CL\")\n",
    "VX_auc = all_models(\"predict_result3\",\"VX\")\n",
    "RTY_auc = all_models(\"predict_result2\",\"RTY\")\n",
    "\n",
    "total_auc2 = pd.concat([YM_auc, NQ_auc, ES_auc, PA_auc, EW_auc, UB_auc, CL_auc, VX_auc, RTY_auc]).reset_index().iloc[:,1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "023439af-f471-4401-853c-02bb392a40d9",
   "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>delta</th>\n",
       "      <th>variable</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>d2.0</td>\n",
       "      <td>4h</td>\n",
       "      <td>0.018358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>d2.2</td>\n",
       "      <td>4h</td>\n",
       "      <td>0.017015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>d2.4</td>\n",
       "      <td>4h</td>\n",
       "      <td>-0.003091</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>d2.6</td>\n",
       "      <td>4h</td>\n",
       "      <td>0.017932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>d2.8</td>\n",
       "      <td>4h</td>\n",
       "      <td>-0.013929</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>457</th>\n",
       "      <td>d3.2</td>\n",
       "      <td>4h</td>\n",
       "      <td>-0.009850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>458</th>\n",
       "      <td>d3.4</td>\n",
       "      <td>4h</td>\n",
       "      <td>-0.004658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>459</th>\n",
       "      <td>d3.6</td>\n",
       "      <td>4h</td>\n",
       "      <td>0.005118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>460</th>\n",
       "      <td>d3.8</td>\n",
       "      <td>4h</td>\n",
       "      <td>-0.005509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>461</th>\n",
       "      <td>d4.0</td>\n",
       "      <td>4h</td>\n",
       "      <td>0.000400</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>462 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    delta variable     value\n",
       "0    d2.0       4h  0.018358\n",
       "1    d2.2       4h  0.017015\n",
       "2    d2.4       4h -0.003091\n",
       "3    d2.6       4h  0.017932\n",
       "4    d2.8       4h -0.013929\n",
       "..    ...      ...       ...\n",
       "457  d3.2       4h -0.009850\n",
       "458  d3.4       4h -0.004658\n",
       "459  d3.6       4h  0.005118\n",
       "460  d3.8       4h -0.005509\n",
       "461  d4.0       4h  0.000400\n",
       "\n",
       "[462 rows x 3 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total_auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "4bf300d8-7fb1-46ef-8b1b-4b48b72bae40",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Label이 몇개나 변한지..?실제로?\n",
    "target_label_list = ['Target_Label', 'Target_2.0_Label', 'Target_2.2_Label',\n",
    "       'Target_2.4_Label', 'Target_2.6_Label', 'Target_2.8_Label',\n",
    "       'Target_3.0_Label', 'Target_3.2_Label', 'Target_3.4_Label',\n",
    "       'Target_3.6_Label', 'Target_3.8_Label', 'Target_4.0_Label']\n",
    "\n",
    "\n",
    "btc_4h   = pd.read_csv(\"../prepro/BTC_4h.csv\")[target_label_list]\n",
    "eth_4h   = pd.read_csv(\"../prepro/ETH_4h.csv\")[target_label_list]\n",
    "doge_4h   = pd.read_csv(\"../prepro/DOGE_4h.csv\")[target_label_list]\n",
    "sol_4h   = pd.read_csv(\"../prepro/SOL_4h.csv\")[target_label_list]\n",
    "bnb_4h   = pd.read_csv(\"../prepro/BNB_4h.csv\")[target_label_list]\n",
    "ada_4h   = pd.read_csv(\"../prepro/ADA_4h.csv\")[target_label_list]\n",
    "xrp_4h  = pd.read_csv(\"../prepro/XRP_4h.csv\")[target_label_list]\n",
    "\n",
    "btc_label_change_4h  =(  btc_4h.sum()[1:] - btc_4h.sum()[0]   ).abs()\n",
    "eth_label_change_4h  =(  eth_4h.sum()[1:] - eth_4h.sum()[0]   ).abs()\n",
    "sol_label_change_4h  =(  sol_4h.sum()[1:] - sol_4h.sum()[0] ).abs()\n",
    "bnb_label_change_4h  =(  bnb_4h.sum()[1:] - bnb_4h.sum()[0] ).abs()\n",
    "ada_label_change_4h  =(  ada_4h.sum()[1:] - ada_4h.sum()[0] ).abs()\n",
    "xrp_label_change_4h  =(  xrp_4h.sum()[1:] - xrp_4h.sum()[0] ).abs()\n",
    "dog_label_change_4h = ( doge_4h.sum()[1:] - doge_4h.sum()[0] ).abs()\n",
    "\n",
    "\n",
    "all_change_4h = pd.concat([\n",
    "pd.DataFrame(btc_label_change_4h ).T,\n",
    "pd.DataFrame(eth_label_change_4h ).T,\n",
    "pd.DataFrame(sol_label_change_4h ).T,\n",
    "pd.DataFrame(bnb_label_change_4h ).T,\n",
    "pd.DataFrame(ada_label_change_4h ).T,\n",
    "pd.DataFrame(xrp_label_change_4h ).T,\n",
    "pd.DataFrame(dog_label_change_4h ).T,\n",
    "\n",
    "])\n",
    "\n",
    "all_change_4h.columns = ['d0.40','d0.44', 'd0.48', 'd0.52', 'd0.56', 'd0.60', 'd0.64', 'd0.68', 'd0.72', 'd0.76', 'd0.80']\n",
    "all_change_4h = all_change_4h.melt()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "200095f3-f4bd-491b-9b99-ed3a0929aa4f",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "### Label이 몇개나 변한지..?실제로?\n",
    "target_label_list = ['Target_Label', 'Target_2.0_Label', 'Target_2.2_Label',\n",
    "       'Target_2.4_Label', 'Target_2.6_Label', 'Target_2.8_Label',\n",
    "       'Target_3.0_Label', 'Target_3.2_Label', 'Target_3.4_Label',\n",
    "       'Target_3.6_Label', 'Target_3.8_Label', 'Target_4.0_Label']\n",
    "\n",
    "\n",
    "btc_4h   = pd.read_csv(\"../prepro/YM_4h.csv\")[target_label_list]\n",
    "eth_4h   = pd.read_csv(\"../prepro/NQ_4h.csv\")[target_label_list]\n",
    "doge_4h   = pd.read_csv(\"../prepro/ES_4h.csv\")[target_label_list]\n",
    "sol_4h   = pd.read_csv(\"../prepro/PA_4h.csv\")[target_label_list]\n",
    "bnb_4h   = pd.read_csv(\"../prepro/EW_4h.csv\")[target_label_list]\n",
    "ada_4h   = pd.read_csv(\"../prepro/UB_4h.csv\")[target_label_list]\n",
    "xrp_4h  = pd.read_csv(\"../prepro/CL_4h.csv\")[target_label_list]\n",
    "vx_4h  = pd.read_csv(\"../prepro/VX_4h.csv\")[target_label_list]\n",
    "rty_4h  = pd.read_csv(\"../prepro/RTY_4h.csv\")[target_label_list]\n",
    "\n",
    "btc_label_change_4h  =(  btc_4h.sum()[1:] - btc_4h.sum()[0]   ).abs()\n",
    "eth_label_change_4h  =(  eth_4h.sum()[1:] - eth_4h.sum()[0]   ).abs()\n",
    "sol_label_change_4h  =(  sol_4h.sum()[1:] - sol_4h.sum()[0] ).abs()\n",
    "bnb_label_change_4h  =(  bnb_4h.sum()[1:] - bnb_4h.sum()[0] ).abs()\n",
    "ada_label_change_4h  =(  ada_4h.sum()[1:] - ada_4h.sum()[0] ).abs()\n",
    "xrp_label_change_4h  =(  xrp_4h.sum()[1:] - xrp_4h.sum()[0] ).abs()\n",
    "dog_label_change_4h = ( doge_4h.sum()[1:] - doge_4h.sum()[0] ).abs()\n",
    "vx_label_change_4h  =(  vx_4h.sum()[1:] - vx_4h.sum()[0] ).abs()\n",
    "rty_label_change_4h = ( rty_4h.sum()[1:] - rty_4h.sum()[0] ).abs()\n",
    "\n",
    "\n",
    "all_stock_change_4h = pd.concat([\n",
    "pd.DataFrame(btc_label_change_4h ).T,\n",
    "pd.DataFrame(eth_label_change_4h ).T,\n",
    "pd.DataFrame(sol_label_change_4h ).T,\n",
    "pd.DataFrame(bnb_label_change_4h ).T,\n",
    "pd.DataFrame(ada_label_change_4h ).T,\n",
    "pd.DataFrame(xrp_label_change_4h ).T,\n",
    "pd.DataFrame(dog_label_change_4h ).T,\n",
    "pd.DataFrame(vx_label_change_4h ).T,\n",
    "pd.DataFrame(rty_label_change_4h ).T,\n",
    "])\n",
    "\n",
    "all_stock_change_4h.columns = ['d0.40','d0.44', 'd0.48', 'd0.52', 'd0.56', 'd0.60', 'd0.64', 'd0.68', 'd0.72', 'd0.76', 'd0.80']\n",
    "all_stock_change_4h = all_stock_change_4h.melt()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "5d52d365-db4e-4c42-a8b4-a54679bb532b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x500 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style({\"style\":\"white\",'font.family':'serif', 'font.serif':'Times New Roman', \"font_scale\" : 1.7})\n",
    "fig, axes = plt.subplots(2, 2, figsize=(12, 5))\n",
    "plt.rcParams[\"font.family\"] = \"Times New Roman\"\n",
    "\n",
    "\n",
    "g = sns.boxplot(x = \"delta\", y=\"value\", data = total_auc2[total_auc2[\"variable\"] == \"4h\"], ax = axes[1, 0], palette = \"gist_gray_r\", showfliers = True)\n",
    "g.axhline(0.0, linestyle = \"--\", linewidth=3, color = \"b\")\n",
    "g.set_title(\"\")\n",
    "g.legend([],[], frameon=False)\n",
    "g.set_ylabel(\"AUC difference\")\n",
    "g.set_xlabel(\"maximum tolerance level ($\\delta$)\")\n",
    "g.set_xticklabels(['0.50','0.60', '0.70', '0.80', '0.90', '1.00', '1.10', '1.20', '1.30', '1.40', '1.50'])\n",
    "\n",
    "\n",
    "\n",
    "g = sns.boxplot(x = \"variable\", y=\"value\", data = all_stock_change_4h, ax = axes[0, 0], palette = \"gist_gray_r\", showfliers = True)\n",
    "#g.axhline(0.0, linestyle = \"--\", linewidth=3, color = \"b\")\n",
    "#g.axhline(melt_4h.iloc[1, 2], linestyle = \":\", linewidth=3, color = \"orange\")\n",
    "g.legend([],[], frameon=False)\n",
    "g.set_title(\"U.S. Futures\")\n",
    "g.set_ylabel(\"Label difference\")\n",
    "g.set_xlabel(\"\")\n",
    "g.set_xticklabels(['0.50','0.60', '0.70', '0.80', '0.90', '1.00', '1.10', '1.20', '1.30', '1.40', '1.50'])\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "g = sns.boxplot(x = \"delta\", y=\"value\", data = total_auc[total_auc[\"variable\"] == \"4h\"], ax = axes[1, 1], palette = \"gist_gray_r\", showfliers = True)\n",
    "g.axhline(0.0, linestyle = \"--\", linewidth=3, color = \"b\")\n",
    "#g.axhline(melt_4h.iloc[1, 2], linestyle = \":\", linewidth=3, color = \"orange\")\n",
    "g.legend([],[], frameon=False)\n",
    "g.set_ylabel(\"\")\n",
    "g.set_xlabel(\"maximum tolerance level ($\\delta$)\")\n",
    "g.set_xticklabels(['1.00','1.20', '1.40', '1.60', '1.80', '2.00', '2.20', '2.40', '2.60', '2.80', '3.00'])\n",
    "\n",
    "g = sns.boxplot(x = \"variable\", y=\"value\", data = all_change_4h, ax = axes[0, 1], palette = \"gist_gray_r\", showfliers = True)\n",
    "#g.axhline(melt_4h.iloc[1, 2], linestyle = \":\", linewidth=3, color = \"orange\")\n",
    "g.legend([],[], frameon=False)\n",
    "g.set_title(\"Cryptocurrency\")\n",
    "g.set_ylabel(\"\")\n",
    "g.set_xlabel(\"\")\n",
    "g.set_xticklabels(['1.00','1.20', '1.40', '1.60', '1.80', '2.00', '2.20', '2.40', '2.60', '2.80', '3.00'])\n",
    "\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"Result_of_model_evaluation.pdf\")\n",
    "plt.savefig(\"Result_of_model_evaluation.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "171530d1-941c-4e81-b6c8-d575deee2eb7",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c6b55f94-25a5-4b59-909f-c8066fa89461",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ad503218-4bf9-46d6-a4fb-8fd3df2263c1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1ddc9818-0d9e-47d9-8018-70e7e0a7c7a3",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "35206170-5a46-4cfd-9ef0-c32c57d99328",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "bigcon",
   "language": "python",
   "name": "bigcon"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}