{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1338\n",
      "1044\n"
     ]
    }
   ],
   "source": [
    "\n",
    "import pandas as pd\n",
    "import json\n",
    "\n",
    "# Replace 'your_file_name.csv' with the path to your CSV file\n",
    "file_name = 'all_pairvote_log_wgpt.csv'\n",
    "\n",
    "# Load the CSV file into a DataFrame\n",
    "df = pd.read_csv(file_name)\n",
    "\n",
    "# Define a function to parse JSON data\n",
    "def parse_json(data):\n",
    "    try:\n",
    "        # Parse the JSON data\n",
    "        return json.loads(data)\n",
    "    except ValueError as e:\n",
    "        # Return None or an empty dictionary if the data cannot be parsed\n",
    "        return None\n",
    "\n",
    "# Apply the parse_json function to the 'models' and 'states' columns\n",
    "df['models'] = df['models'].apply(parse_json)\n",
    "df['states'] = df['states'].apply(parse_json)\n",
    "# row[\"states\"][0]['messages'][0][1]\n",
    "\n",
    "# Now df contains the parsed JSON data in the 'models' and 'states' columns\n",
    "# print(df.head())\n",
    "print(len(df))\n",
    "# filter_vote_df = df[df[\"gpt_vote\"].isin([\"leftvote\", \"rightvote\"])]#, \"tievote\", \"bothbad_vote\"\n",
    "# \\#1\n",
    "filter_vote_df = df[df[\"gpt_vote\"].isin([\"leftvote\", \"rightvote\", \"tievote\", \"bothbad_vote\"])]\n",
    "# \\#2\n",
    "# filter_vote_df = df\n",
    "filter_vote_df.loc[~filter_vote_df[\"gpt_vote\"].isin([\"leftvote\", \"rightvote\"]), \"gpt_vote\"] = \"tie\"\n",
    "filter_vote_df.loc[~filter_vote_df[\"type\"].isin([\"leftvote\", \"rightvote\"]), \"type\"] = \"tie\"\n",
    "# \\#3\n",
    "#[df[\"gpt_vote\"].isin([\"leftvote\", \"rightvote\"]) & df[\"type\"].isin([\"leftvote\", \"rightvote\"])]\n",
    "filtered_df = filter_vote_df[filter_vote_df[\"states\"].apply(lambda x: len(x[0]['messages'][0][1]) > 10)]\n",
    "print(len(filtered_df))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix:\n",
      "[[300  61  34]\n",
      " [102 269  27]\n",
      " [ 99 111  41]]\n",
      "\n",
      "Accuracy: 0.5842911877394636\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score\n",
    "import pandas as pd\n",
    "\n",
    "# Assuming df is your DataFrame\n",
    "\n",
    "# True labels\n",
    "y_true = filtered_df[\"type\"]\n",
    "\n",
    "# Predictions\n",
    "y_pred = filtered_df[\"gpt_vote\"]\n",
    "\n",
    "# Compute the confusion matrix\n",
    "# conf_matrix = confusion_matrix(y_true, y_pred, labels=[\"leftvote\", \"rightvote\", \"tievote\", \"bothbad_vote\"])\n",
    "conf_matrix = confusion_matrix(y_true, y_pred, labels=[\"leftvote\", \"rightvote\", \"tie\"])\n",
    "\n",
    "# Compute the accuracy\n",
    "accuracy = accuracy_score(y_true, y_pred)\n",
    "\n",
    "print(\"Confusion Matrix:\")\n",
    "print(conf_matrix)\n",
    "\n",
    "print(\"\\nAccuracy:\", accuracy)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import pandas as pd\n",
    "\n",
    "# Assuming df is your DataFrame\n",
    "\n",
    "# True labels and predictions\n",
    "y_true = filtered_df[\"type\"]\n",
    "y_pred = filtered_df[\"gpt_vote\"]\n",
    "\n",
    "# Compute the confusion matrix\n",
    "conf_matrix = confusion_matrix(y_true, y_pred, labels=[\"leftvote\", \"rightvote\", \"tievote\", \"bothbad_vote\"])\n",
    "\n",
    "# Create a pandas DataFrame from the confusion matrix\n",
    "conf_matrix_df = pd.DataFrame(conf_matrix, index=[\"leftvote\", \"rightvote\", \"tievote\", \"bothbad_vote\"], columns=[\"leftvote\", \"rightvote\", \"tievote\", \"bothbad_vote\"])\n",
    "\n",
    "# Plotting the heatmap\n",
    "plt.figure(figsize=(10, 7))\n",
    "sns.heatmap(conf_matrix_df, annot=True, fmt=\"d\", cmap=\"Blues\", cbar=False)\n",
    "plt.title(\"Arena Human vs GPT-4V Confusion Matrix\")\n",
    "plt.xlabel(\"GPT-4V Vote\")\n",
    "plt.ylabel(\"Arena Human Vote\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.5842911877394636\n",
      "F1 Score (Macro): 0.514392348541452\n",
      "F1 Score (Micro): 0.5842911877394636\n",
      "F1 Score (Weighted): 0.5536668839130223\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score, f1_score\n",
    "\n",
    "# Assuming df is your DataFrame and it contains 'type' as true labels and 'gpt_vote' as predictions\n",
    "y_true = filtered_df['type']\n",
    "y_pred = filtered_df['gpt_vote']\n",
    "\n",
    "# Calculate accuracy\n",
    "accuracy = accuracy_score(y_true, y_pred)\n",
    "print(f'Accuracy: {accuracy}')\n",
    "\n",
    "# Calculate F1 score, here using 'macro' average to treat all classes equally\n",
    "f1 = f1_score(y_true, y_pred, average='macro')\n",
    "print(f'F1 Score (Macro): {f1}')\n",
    "\n",
    "# If you want to calculate F1 score with other averages, for example 'micro' or 'weighted', you can do:\n",
    "f1_micro = f1_score(y_true, y_pred, average='micro')\n",
    "print(f'F1 Score (Micro): {f1_micro}')\n",
    "\n",
    "f1_weighted = f1_score(y_true, y_pred, average='weighted')\n",
    "print(f'F1 Score (Weighted): {f1_weighted}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cohen's Kappa Score: 0.3442144615665177\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import cohen_kappa_score\n",
    "\n",
    "# Assuming df is your DataFrame and it contains 'type' as true labels and 'gpt_vote' as predictions\n",
    "y_true = filtered_df['type']\n",
    "y_pred = filtered_df['gpt_vote']\n",
    "\n",
    "# Calculate Cohen's Kappa score\n",
    "kappa = cohen_kappa_score(y_true, y_pred)\n",
    "print(f'Cohen\\'s Kappa Score: {kappa}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score: 0.5842911877394636\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "accuracy = accuracy_score(y_true, y_pred)\n",
    "print(f'Accuracy Score: {accuracy}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pearson Correlation Coefficient: 0.2880096104357029\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# Assuming filtered_df is your DataFrame and it contains 'type' and 'gpt_vote' columns\n",
    "# Convert 'type' and 'gpt_vote' to categorical codes\n",
    "filtered_df['type_int'] = pd.factorize(filtered_df['type'])[0]\n",
    "filtered_df['gpt_vote_int'] = pd.factorize(filtered_df['gpt_vote'])[0]\n",
    "\n",
    "# Now you can calculate Pearson correlation between these new integer columns\n",
    "pearson_correlation = filtered_df['type_int'].corr(filtered_df['gpt_vote_int'])\n",
    "print(f'Pearson Correlation Coefficient: {pearson_correlation}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "otask",
   "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": "3.10.13"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}