{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Generator"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Create a list of parties that participate in elections. It is better if the parties' names have a fictional nature, but somehow relates to your country's politics:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "general_list_of_parties = ['Harmony Party', 'Red Party', 'Justice and Truth', 'Adventure Alliance', 'Animal Friends Party', 'Go Greens!', 'Yo-ho-ho Pirate Party']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Generate ballots of all formats. We used 2 formats: ranked ballots with party names and positional ranking with numbers:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13000"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random.seed(42)\n",
    "\n",
    "# correct ballots\n",
    "import random\n",
    "import numpy as np\n",
    "# here you can assign some distribution to party votes, e.g. based on the last elections. \n",
    "# If you want to underline the ambiguity about the best system for counting votes, it is better to make probabilities almost equal.\n",
    "prob_weights = np.array([0.20, 0.20, 0.20, 0.19, 0.19, 0.01, 0.01])\n",
    "\n",
    "# generating votes_names\n",
    "votes = []\n",
    "for i in range(3,8):\n",
    "    for k in range(random.randint(1000,1500)):\n",
    "        votes.append([str(i) for i in list(np.random.choice(general_list_of_parties, size=i, replace=False, p=prob_weights))])\n",
    "\n",
    "# generating votes_ranks\n",
    "votes2 = []\n",
    "for i in range(2,8):\n",
    "    for k in range(random.randint(1000,1500)):\n",
    "        comb = [str(i) for i in list(np.random.choice(general_list_of_parties, size=i, replace=False, p=prob_weights))]\n",
    "        newlist = []\n",
    "        for party in general_list_of_parties:\n",
    "            if party in comb:\n",
    "                newlist.append(comb.index(party)+1)\n",
    "            else:\n",
    "                newlist.append(0)\n",
    "        votes2.append(newlist)\n",
    "\n",
    "votes += votes2\n",
    "random.shuffle(votes)\n",
    "\n",
    "len(votes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. Generate spoiled ballots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "101"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# incorrect ballots\n",
    "general_list_of_parties_d = ['Go Grass!', 'Advertisement Party', 'Pink Party', 'Justice and Teeth', 'SDP', 'Party like a Racoon', 'Pickle Ricks', 'Princess Party']\n",
    "\n",
    "names_d = []\n",
    "for i in range(3,8):\n",
    "    for k in range(random.randint(5, 25)):\n",
    "        names_d.append([str(i) for i in list(np.random.choice(general_list_of_parties, size=8, replace=True, p=[1/7] * 7))])\n",
    "\n",
    "len(names_d)\n",
    "\n",
    "ranks_d = []\n",
    "\n",
    "for k in range(random.randint(20, 40)):\n",
    "    sample = list(np.random.choice(general_list_of_parties, size=7, replace=False, p=prob_weights))\n",
    "    comb = [str(i) for i in sample]\n",
    "    newlist = []\n",
    "    for party in general_list_of_parties:\n",
    "        newlist.append(comb.index(party)+2)\n",
    "    ranks_d.append(newlist)\n",
    "\n",
    "for k in range(random.randint(20, 40)):\n",
    "    sample = list(np.random.choice(general_list_of_parties, size=random.randint(3,7), replace=False, p=prob_weights))\n",
    "    comb = [str(i) for i in sample]\n",
    "    newlist = []\n",
    "    for party in general_list_of_parties:\n",
    "        if party in comb:\n",
    "            newlist.append(0)\n",
    "        else:\n",
    "            newlist.append(0)\n",
    "    ranks_d.append(newlist)\n",
    "\n",
    "for k in range(random.randint(20, 40)):\n",
    "    sample = list(np.random.choice(general_list_of_parties, size=random.randint(3,7), replace=False, p=prob_weights))\n",
    "    comb = [str(i) for i in sample]\n",
    "    newlist = []\n",
    "    for party in general_list_of_parties:\n",
    "        if party in comb:\n",
    "            newlist.append(0)\n",
    "        else:\n",
    "            newlist.append(0)\n",
    "    newlist.append(random.randint(0,1))\n",
    "    ranks_d.append(newlist)\n",
    "\n",
    "len(ranks_d)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "4. Bind and shuffle all the ballots, save as `.csv` file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13159"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "votes = votes + names_d + ranks_d\n",
    "random.shuffle(votes)\n",
    "len(votes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "# saving data\n",
    "import csv\n",
    "with open(\"votes.csv\", \"w\") as f:\n",
    "    wr = csv.writer(f)\n",
    "    wr.writerows(votes)"
   ]
  }
 ],
 "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": "3.10.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}