siimee/backend/db/data-generator/generate.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Scoring: ((1, 1), (7, 7))\n",
      "1515\n",
      "\n",
      "Scoring: ((2, 1), (2, 2))\n",
      "9000\n",
      "\n",
      "Scoring: ((2, 7), (6, 3))\n",
      "4343\n",
      "\n",
      "Scoring: ((6, 3), (1, 2))\n",
      "4901\n",
      "\n",
      "Scoring: ((5, 6), (4, 6))\n",
      "9000\n",
      "\n",
      "Scoring: ((1, 2), (3, 3))\n",
      "7764\n",
      "\n",
      "Scoring: ((1, 1), (1, 1))\n",
      "10000\n"
     ]
    }
   ],
   "source": [
    "import json\n",
    "import copy\n",
    "from itertools import product\n",
    "from scipy import spatial\n",
    "import math\n",
    "import numpy\n",
    "\n",
    " \n",
    "f = open('./config.json')\n",
    "from_json_config = json.load(f)\n",
    "config = {}\n",
    "config[\"all_possible_responses\"] = from_json_config[\"scoreVals\"]\n",
    "config[\"magic\"] = from_json_config[\"magic\"]\n",
    "config[\"aspect_size\"] = 2\n",
    "config[\"write\"]= True\n",
    "config[\"file\"] = \"../generated/prescore_matrix.json\"\n",
    "config[\"delimiter\"] = \"-\"\n",
    "config[\"version\"] = \"0.1.0\"\n",
    "# max distance = distance between top left and bottom right indices in matrix\n",
    "config[\"max_euclidean_distance\"] =  math.sqrt(\n",
    "    math.pow(len(config[\"all_possible_responses\"])-1, 2) +\n",
    "    math.pow(len(config[\"all_possible_responses\"])-1, 2)\n",
    "    ) \n",
    "f.close()\n",
    "\n",
    "def createPermutations(possibilities, size):\n",
    "    return tuple(product(possibilities, repeat=size))\n",
    "\n",
    "\n",
    "def getAspectFromSurveys(survey_a, survey_b, size):\n",
    "    if (survey_a.__len__() < size | survey_b.__len__() < size):\n",
    "        raise Exception(\"Surveys must both contain more items than size\")\n",
    "\n",
    "    def store(survey, length):\n",
    "        col = []\n",
    "        for i in range(size):\n",
    "            val = survey.pop(0)\n",
    "            col.append(val)\n",
    "        return col\n",
    "\n",
    "    # Take the first <size> elements from the list\n",
    "    col_a = tuple(store(survey_a, size))\n",
    "    col_b = tuple(store(survey_b, size))\n",
    "\n",
    "    if (col_a.__len__() != size | col_b.__len__() != size ):\n",
    "        raise Exception(\"No aspect values found in survey\")\n",
    "\n",
    "    return (col_a, col_b)\n",
    "\n",
    "\n",
    "def scoreAspect(aspect_ab):\n",
    "    a = aspect_ab[0]\n",
    "    b = aspect_ab[1]\n",
    "    return (1 - spatial.distance.cosine(a,b)) * config[\"magic\"]\n",
    "\n",
    "def euclideanDistance(aspect_ab):\n",
    "    a = numpy.array(aspect_ab[0])\n",
    "    b = numpy.array(aspect_ab[1])\n",
    "    return numpy.linalg.norm(a - b)\n",
    "\n",
    "def prescore_matrix_from(vals):\n",
    "    m = {}\n",
    "    for val in vals:\n",
    "        m[val] = []\n",
    "        for other_val in vals:\n",
    "            distance = euclideanDistance((val,other_val))\n",
    "            score = (config[\"max_euclidean_distance\"] - distance) * config[\"magic\"] + 1515\n",
    "            # score = scoreAspect((val, other_val))\n",
    "            adjusted_score = round(score)\n",
    "            m[val].append(adjusted_score)\n",
    "    return m\n",
    "\n",
    "\n",
    "def lookup_prescore_in(score_matrix, vals, aspect_ab):\n",
    "    print(\"\\nScoring:\", aspect_ab)\n",
    "    aspect_a, aspect_b = aspect_ab\n",
    "    # Look-up using the index because\n",
    "    # \n",
    "    pos_b = vals.index(aspect_b)\n",
    "    return score_matrix[aspect_a][pos_b]\n",
    "\n",
    "\n",
    "# !: Mutates your input\n",
    "def score_aspect(input_a, input_b, score_matrix, vals):\n",
    "    aspect_ab = getAspectFromSurveys(input_a, input_b, config[\"aspect_size\"])\n",
    "    return lookup_prescore_in(score_matrix, vals, aspect_ab)\n",
    "\n",
    "\n",
    "def run():\n",
    "    # Set the keys for the look-up\n",
    "    xy_axis_vals = createPermutations(config[\"all_possible_responses\"], config[\"aspect_size\"])\n",
    "    m = prescore_matrix_from(xy_axis_vals)\n",
    "\n",
    "    # Example:\n",
    "    res = config[\"all_possible_responses\"]\n",
    "    input_a = [\n",
    "        res[0], res[0], # One aspect\n",
    "        res[1], res[0],\n",
    "        res[1], res[6],\n",
    "        res[5], res[2],\n",
    "        res[4], res[5],\n",
    "        res[0], res[1],\n",
    "        res[0], res[0],\n",
    "    ]\n",
    "    input_b = [\n",
    "        res[6], res[6], # One aspect\n",
    "        res[1], res[1],\n",
    "        res[5], res[2],\n",
    "        res[0], res[1],\n",
    "        res[3], res[5],\n",
    "        res[2], res[2],\n",
    "        res[0], res[0],\n",
    "    ]\n",
    "    for i in range(round(input_a.__len__() / 2)):\n",
    "        print(score_aspect(input_a, input_b, m, xy_axis_vals))\n",
    "\n",
    "\n",
    "    if(config[\"write\"] == True):\n",
    "        # Serializing json\n",
    "        str_m = {}\n",
    "        for key, values in m.items():\n",
    "            delimiter = config[\"delimiter\"]\n",
    "            str_key = delimiter.join([str(v) for v in key])\n",
    "            str_m[str_key] = [int(val) for val in values]\n",
    "            str_m[\"_config\"] = config\n",
    "        json_object = json.dumps(str_m, indent = 4)\n",
    "        with open(config[\"file\"], \"w\") as file:\n",
    "            # write to file\n",
    "            file.write(json_object)\n",
    "run()\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.6 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": "3.10.6"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "b0fa6594d8f4cbf19f97940f81e996739fb7646882a419484c72d19e05852a7e"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}