{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## FUNKCJE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def inertia_list(all_doc):\n",
    "    list_inter = []\n",
    "    K_max = int(len(all_doc)/2)\n",
    "    while K_max > 100:\n",
    "        K_max = int(K_max/2)\n",
    "    K = range(1,K_max)\n",
    "    for k in K:\n",
    "        FitMean = KMeans(n_clusters=k).fit(doc_vectors)\n",
    "        list_inter.append(FitMean.inertia_)\n",
    "    return list_inter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def BestK(list_inter):\n",
    "    position = -10\n",
    "    for i in range(0, len(list_inter)-1):\n",
    "        if (int(list_inter[i]) == (int(list_inter[i+1]))):\n",
    "            position = i\n",
    "    if position == -10 :\n",
    "        position = len(list_inter)-1\n",
    "    return position"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PLIK DEV-0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "infile = open('dev-0/in.tsv', 'r', encoding=\"utf-8\")\n",
    "outfile = open(\"dev-0/out.tsv\", \"w\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_doc = infile.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorizer = TfidfVectorizer()\n",
    "doc_vectors = vectorizer.fit_transform(all_doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_inter = inertia_list(all_doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "position = BestK(list_inter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "FitMean = KMeans(n_clusters=position).fit_predict(doc_vectors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "for x in FitMean:\n",
    "    outfile.write(str(x) + '\\n')\n",
    "infile.close()\n",
    "outfile.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PLIK TEST-A"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "infile = open('test-A/in.tsv', 'r', encoding=\"utf-8\")\n",
    "outfile = open(\"test-A/out.tsv\", \"w\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_doc = infile.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorizer = TfidfVectorizer()\n",
    "doc_vectors = vectorizer.fit_transform(all_doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_inter = inertia_list(all_doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "position = BestK(list_inter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "FitMean = KMeans(n_clusters=position).fit_predict(doc_vectors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "for x in FitMean:\n",
    "    outfile.write(str(x) + '\\n')\n",
    "infile.close()\n",
    "outfile.close()"
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}