{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "import lzma"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = []\n",
    "\n",
    "with lzma.open('train/in.tsv.xz') as f:\n",
    "    for line in f:\n",
    "        X.append(line.decode('utf-8'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "Y = []\n",
    "\n",
    "with open('train/expected.tsv') as f:\n",
    "    for line in f:\n",
    "        txt = line\n",
    "        txt = txt.replace('\\n', '')\n",
    "        Y.append(txt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "\n",
    "vectorizer = TfidfVectorizer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "textVectors = vectorizer.fit_transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[0., 0., 0., ..., 0., 0., 0.]])"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "textVectors[0].todense()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.naive_bayes import BernoulliNB\n",
    "import numpy as np\n",
    "\n",
    "trainY = np.array(Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "BernoulliNB()"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bernoulli = BernoulliNB()\n",
    "bernoulli.fit(textVectors, trainY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "import csv\n",
    "\n",
    "testX = []\n",
    "\n",
    "with open('dev-0/in.tsv', encoding='utf8') as f:\n",
    "    for line in f:\n",
    "        testX.append(line)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "testX = vectorizer.transform(testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictedY = bernoulli.predict(testX)\n",
    "\n",
    "with open('dev-0/out.tsv', 'w', newline='') as f:\n",
    "    writer = csv.writer(f)\n",
    "    writer.writerows(predictedY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "expectedY = []\n",
    "with open('dev-0/expected.tsv') as f:\n",
    "    for line in f:\n",
    "        txt = line\n",
    "        txt = txt.replace('\\n', '')\n",
    "        expectedY.append(txt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6577260876531162"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bernoulli.score(testX, expectedY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score for dev-1: 0.6406778795193032\n"
     ]
    }
   ],
   "source": [
    "# dev-1\n",
    "testX = []\n",
    "\n",
    "with open('dev-1/in.tsv', encoding='utf8') as f:\n",
    "    for line in f:\n",
    "        testX.append(line)\n",
    "\n",
    "testX = vectorizer.transform(testX)\n",
    "\n",
    "predictedY = bernoulli.predict(testX)\n",
    "\n",
    "expectedY = []\n",
    "\n",
    "with open('dev-1/expected.tsv') as f:\n",
    "    for line in f:\n",
    "        expectedY.append(line.replace('\\n', ''))\n",
    "\n",
    "print('Score for dev-1:', bernoulli.score(testX, expectedY))\n",
    "\n",
    "with open('dev-1/out.tsv', 'w', newline='') as f:\n",
    "    writer = csv.writer(f)\n",
    "    writer.writerows(predictedY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "# test-A\n",
    "testX = []\n",
    "\n",
    "with open('test-A/in.tsv', encoding='utf8') as f:\n",
    "    for line in f:\n",
    "        testX.append(line)\n",
    "\n",
    "testX = vectorizer.transform(testX)\n",
    "\n",
    "predictedY = bernoulli.predict(testX)\n",
    "\n",
    "with open('test-A/out.tsv', 'w', newline='') as f:\n",
    "    writer = csv.writer(f)\n",
    "    writer.writerows(predictedY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "3a5b3979b9a2fc2c8e649de363a592bbf5a2c9da164843b1adb5b45661722ad0"
  },
  "kernelspec": {
   "display_name": "Python 3.8.10 64-bit",
   "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": "3.8.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}