{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c6cadd34",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "import glob"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4ce0c480",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Filenames :\n",
      "../../data/processed/uc-berkeley-measuring-hate-speech-negative-translated.csv\n",
      "../../data/processed/caa_negative.csv\n",
      "../../data/processed/caa_positive.csv\n",
      "../../data/processed/ftr_dataset_negative.csv\n",
      "../../data/processed/mlma_negative.csv\n",
      "../../data/processed/merged_datasets.csv\n",
      "../../data/processed/mlma_positive.csv\n",
      "../../data/processed/ftr_dataset_positive.csv\n",
      "../../data/processed/uc-berkeley-measuring-hate-speech-positive-translated.csv\n",
      "\n",
      "Columns:\n",
      "['Unnamed: 2']\n",
      "['tweet' 'sentiment' 'directness' 'annotator_sentiment' 'target' 'group']\n",
      "['tweet' 'label' 'tweet_clean']\n",
      "['translated' 'hate_speech_score' 'translated_clean' 'label']\n",
      "['translated' 'hate_speech_score' 'translated_clean' 'label']\n",
      "['Unnamed: 2']\n",
      "['tweet' 'label' 'tweet_clean']\n",
      "['tweet' 'sentiment' 'directness' 'annotator_sentiment' 'target' 'group']\n",
      "['text']\n"
     ]
    }
   ],
   "source": [
    "input_dir = \"../../data/processed/\"\n",
    "output_dir = \"../../data/processed/\"\n",
    "\n",
    "filenames = glob.glob(input_dir + \"*\")\n",
    "\n",
    "print(\"Filenames :\")\n",
    "pos, neg = [], []\n",
    "for filename in filenames:\n",
    "    print(filename)\n",
    "    if \"positive\" in filename:\n",
    "        pos.append(pd.read_csv(filename, index_col=[0]))\n",
    "    else:\n",
    "        neg.append(pd.read_csv(filename, index_col=[0]))\n",
    "        \n",
    "print()\n",
    "print(\"Columns:\")\n",
    "for pos_df in pos:\n",
    "    print(pos_df.columns.values)\n",
    "    \n",
    "for neg_df in neg:\n",
    "    print(neg_df.columns.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ef7ee207",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Positive : 5011\n",
      "Negative : 31378\n"
     ]
    }
   ],
   "source": [
    "for pos_df in pos:\n",
    "    if \"Unnamed: 2\" in pos_df.columns:\n",
    "        pos_df.rename({\"Unnamed: 2\" : \"text\"}, inplace=True, axis=1)\n",
    "    elif \"tweet_clean\" in pos_df.columns:\n",
    "        pos_df.rename({\"tweet_clean\" : \"text\"}, inplace=True, axis=1)\n",
    "    elif 'translated_clean' in pos_df.columns:\n",
    "        pos_df.rename({\"translated_clean\" : \"text\"}, inplace=True, axis=1)\n",
    "    elif \"tweet\" in pos_df.columns:\n",
    "        pos_df.rename({\"tweet\" : \"text\"}, inplace=True, axis=1)\n",
    "        \n",
    "for neg_df in neg:\n",
    "    if \"Unnamed: 2\" in neg_df.columns:\n",
    "        neg_df.rename({\"Unnamed: 2\" : \"text\"}, inplace=True, axis=1)\n",
    "    elif \"tweet_clean\" in neg_df.columns:\n",
    "        neg_df.rename({\"tweet_clean\" : \"text\"}, inplace=True, axis=1)\n",
    "    elif 'translated_clean' in neg_df.columns:\n",
    "        neg_df.rename({\"translated_clean\" : \"text\"}, inplace=True, axis=1)\n",
    "    elif \"tweet\" in neg_df.columns:\n",
    "        neg_df.rename({\"tweet\" : \"text\"}, inplace=True, axis=1)\n",
    "        \n",
    "pos = pd.concat(pos)[[\"text\"]]\n",
    "pos[\"label\"] = 1\n",
    "neg = pd.concat(neg)[[\"text\"]]\n",
    "neg[\"label\"] = 0\n",
    "\n",
    "print(\"Positive :\", pos.shape[0])\n",
    "print(\"Negative :\", neg.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "3ec5fa31",
   "metadata": {},
   "outputs": [],
   "source": [
    "merged = pd.concat([pos, neg])\n",
    "merged.to_csv(os.path.join(output_dir, \"merged_datasets.csv\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9fe7d6ab",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fd896f4c",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "sexism_detection",
   "language": "python",
   "name": "sexism_detection"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}