{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "sweet-external",
   "metadata": {},
   "source": [
    "# How to use PCMF package!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "outer-viewer",
   "metadata": {},
   "outputs": [],
   "source": [
    "#import PCMF\n",
    "from PCMF.Positive_CMF import Positive_Collective_Matrix_Factorization\n",
    "#data processing\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dirty-manual",
   "metadata": {},
   "source": [
    "# Make sample dataset(Example data set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "creative-lobby",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[0.24847769, 0.92837696, 0.30059839, ..., 0.12561425, 0.66277199,\n",
       "         0.34027897],\n",
       "        [0.22307267, 0.48443072, 0.57220612, ..., 0.49297108, 0.35689179,\n",
       "         0.41624412],\n",
       "        [0.7361003 , 0.77629826, 0.42066231, ..., 0.97781795, 0.84493988,\n",
       "         0.35037373],\n",
       "        ...,\n",
       "        [0.12404128, 0.0019809 , 0.6390083 , ..., 0.94168058, 0.09853853,\n",
       "         0.38612824],\n",
       "        [0.07181645, 0.56613074, 0.60231026, ..., 0.55167077, 0.71364693,\n",
       "         0.17202591],\n",
       "        [0.88714109, 0.94597517, 0.09069533, ..., 0.71342583, 0.77460753,\n",
       "         0.78309143]]),\n",
       " array([[0.46091998, 0.15893277, 0.50549811, ..., 0.24884235, 0.73122779,\n",
       "         0.98397691],\n",
       "        [0.02287738, 0.3495672 , 0.93856203, ..., 0.98989155, 0.66211367,\n",
       "         0.92033014],\n",
       "        [0.08953751, 0.20408854, 0.79054535, ..., 0.24215255, 0.74023685,\n",
       "         0.32782198],\n",
       "        ...,\n",
       "        [0.1782251 , 0.08667535, 0.64887784, ..., 0.04384896, 0.65464416,\n",
       "         0.27241068],\n",
       "        [0.32575456, 0.10711986, 0.43629966, ..., 0.46911493, 0.61064817,\n",
       "         0.1754898 ],\n",
       "        [0.43208215, 0.29567345, 0.41956385, ..., 0.8987314 , 0.64731738,\n",
       "         0.14023682]]))"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Randomly generate the original matrix.\n",
    "X  = np.random.random_sample((1000, 100))\n",
    "Y = np.random.random_sample((1000, 30))\n",
    "X, Y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "committed-texture",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.59242526, 0.94391661, 0.81555406, ..., 0.45420413, 0.07689032,\n",
       "        0.38228694],\n",
       "       [0.10679816, 0.76132749, 0.8370853 , ..., 0.94615018, 0.65010113,\n",
       "        0.51260314],\n",
       "       [0.69759489, 0.28559725, 0.09380406, ..., 0.24183029, 0.55903652,\n",
       "        0.55620306],\n",
       "       ...,\n",
       "       [0.84851777, 0.8064424 , 0.84641176, ..., 0.81146947, 0.67658812,\n",
       "        0.29319071],\n",
       "       [0.11170263, 0.76870714, 0.83241645, ..., 0.3189608 , 0.04840133,\n",
       "        0.86164286],\n",
       "       [0.07015802, 0.31711469, 0.2241136 , ..., 0.92555989, 0.00483782,\n",
       "        0.80231256]])"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Weight of each element of Y of loss function.\n",
    "wY = np.random.random_sample((1000, 30))\n",
    "wY"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "visible-helen",
   "metadata": {},
   "source": [
    "# Training PCMF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "historical-officer",
   "metadata": {},
   "outputs": [],
   "source": [
    "cmf = Positive_Collective_Matrix_Factorization(X, Y, alpha=0.5, d_hidden=12, lamda=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "posted-cooling",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Info] At time-step 0, loss is 17.701169967651367\n",
      "[Info] At time-step 50, loss is 9.045083999633789\n",
      "[Info] At time-step 100, loss is 5.19536828994751\n",
      "[Info] At time-step 150, loss is 3.2300522327423096\n",
      "[Info] At time-step 200, loss is 1.722853183746338\n",
      "[Info] At time-step 250, loss is 1.0922274589538574\n",
      "[Info] At time-step 300, loss is 1.0090701580047607\n",
      "[Info] At time-step 350, loss is 0.9783576130867004\n",
      "[Info] At time-step 400, loss is 0.9572911858558655\n",
      "[Info] At time-step 450, loss is 0.9419869184494019\n",
      "[Info] At time-step 500, loss is 0.9304507374763489\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(array([[0.00619301, 0.00809157, 0.00185326, ..., 0.00295491, 0.00102456,\n",
       "         0.00220665],\n",
       "        [0.00611927, 0.007062  , 0.00174178, ..., 0.00672322, 0.00129402,\n",
       "         0.00208969],\n",
       "        [0.00177775, 0.00116573, 0.00310689, ..., 0.00278253, 0.00548184,\n",
       "         0.00568578],\n",
       "        ...,\n",
       "        [0.00470339, 0.00284737, 0.00219152, ..., 0.00136558, 0.00203684,\n",
       "         0.00263309],\n",
       "        [0.00225398, 0.00183056, 0.00184434, ..., 0.00957577, 0.00270244,\n",
       "         0.00212347],\n",
       "        [0.00132301, 0.00318799, 0.00137424, ..., 0.00370901, 0.00549211,\n",
       "         0.00241262]], dtype=float32),\n",
       " array([[0.00328673, 0.00322875, 0.00319095, ..., 0.00300826, 0.00441328,\n",
       "         0.00408365],\n",
       "        [0.00308168, 0.00410404, 0.00410875, ..., 0.00431702, 0.00392337,\n",
       "         0.00203434],\n",
       "        [0.00486501, 0.00450918, 0.00258157, ..., 0.00380386, 0.00226068,\n",
       "         0.00241681],\n",
       "        ...,\n",
       "        [0.00401949, 0.00417189, 0.00420266, ..., 0.00427896, 0.00387536,\n",
       "         0.0037397 ],\n",
       "        [0.00444068, 0.00399209, 0.00425376, ..., 0.00415507, 0.00242299,\n",
       "         0.00392819],\n",
       "        [0.0037223 , 0.00423859, 0.00303006, ..., 0.00450027, 0.00434054,\n",
       "         0.00723165]], dtype=float32),\n",
       " array([[0.00203041, 0.00376991, 0.00133253, 0.00786313, 0.00257054,\n",
       "         0.00449617, 0.007386  , 0.0019676 , 0.0023274 , 0.00209709,\n",
       "         0.00200457, 0.00219077, 0.00198193, 0.00706211, 0.00227018,\n",
       "         0.00481099, 0.00537675, 0.00124624, 0.00234203, 0.00193601,\n",
       "         0.00210485, 0.00208073, 0.0052475 , 0.00182663, 0.00518937,\n",
       "         0.00415232, 0.00221185, 0.00352488, 0.00112584, 0.00541137],\n",
       "        [0.00786245, 0.0013524 , 0.00294638, 0.00209544, 0.00152745,\n",
       "         0.00171906, 0.00206297, 0.0036735 , 0.0014902 , 0.00108148,\n",
       "         0.00619554, 0.0020093 , 0.00211902, 0.00214887, 0.00600566,\n",
       "         0.0041916 , 0.00765501, 0.00143352, 0.00230244, 0.01046781,\n",
       "         0.00196616, 0.00523918, 0.00141301, 0.01054175, 0.00335738,\n",
       "         0.00449362, 0.00234817, 0.00410568, 0.00351846, 0.00220969],\n",
       "        [0.00224716, 0.00442828, 0.01085493, 0.00311857, 0.00219517,\n",
       "         0.00283331, 0.00404022, 0.00163152, 0.00256795, 0.00198213,\n",
       "         0.00217025, 0.00572459, 0.00205693, 0.00228126, 0.00520555,\n",
       "         0.00637635, 0.00138925, 0.00335152, 0.0049909 , 0.00351976,\n",
       "         0.00899771, 0.00465523, 0.00293077, 0.00214331, 0.00488522,\n",
       "         0.00404807, 0.0034802 , 0.00184786, 0.00557039, 0.00337978],\n",
       "        [0.00201379, 0.00207763, 0.00268584, 0.00226676, 0.00293873,\n",
       "         0.01008144, 0.00952133, 0.00227971, 0.00218996, 0.0015605 ,\n",
       "         0.00574247, 0.00192397, 0.00221183, 0.00177528, 0.0021055 ,\n",
       "         0.00483836, 0.00147783, 0.00213373, 0.00374119, 0.00227146,\n",
       "         0.00466285, 0.00233669, 0.00374851, 0.0047995 , 0.00202598,\n",
       "         0.0017043 , 0.00354986, 0.00212978, 0.00501988, 0.00422556],\n",
       "        [0.00520168, 0.00253476, 0.00183491, 0.00461178, 0.00474329,\n",
       "         0.00267396, 0.00142287, 0.00229034, 0.00214646, 0.00526973,\n",
       "         0.00213205, 0.00225656, 0.00331753, 0.00302078, 0.002235  ,\n",
       "         0.0021887 , 0.00226603, 0.00292394, 0.00131771, 0.00196172,\n",
       "         0.01076068, 0.00266699, 0.00221681, 0.00203325, 0.00920436,\n",
       "         0.00187069, 0.00193321, 0.00468633, 0.00203348, 0.0088259 ],\n",
       "        [0.00280944, 0.00278756, 0.00181516, 0.0021846 , 0.00521531,\n",
       "         0.00107158, 0.00185494, 0.00305022, 0.00216553, 0.00122412,\n",
       "         0.00184354, 0.00208403, 0.00184739, 0.00166199, 0.00241919,\n",
       "         0.00154003, 0.00603467, 0.00338922, 0.00222716, 0.00118293,\n",
       "         0.00105201, 0.0033551 , 0.00189878, 0.00147722, 0.00898661,\n",
       "         0.00141901, 0.00123493, 0.00385798, 0.00379461, 0.0022931 ],\n",
       "        [0.0031502 , 0.00475225, 0.00865962, 0.00303563, 0.00182524,\n",
       "         0.00310831, 0.00373822, 0.00209216, 0.00199991, 0.00354059,\n",
       "         0.00470833, 0.00584881, 0.00214915, 0.00227718, 0.0012489 ,\n",
       "         0.00503554, 0.00162324, 0.00518536, 0.00220567, 0.00503302,\n",
       "         0.00514321, 0.00171418, 0.00533724, 0.0020934 , 0.00191555,\n",
       "         0.00274799, 0.00538925, 0.00432319, 0.00169857, 0.00504138],\n",
       "        [0.00325267, 0.00193892, 0.00139512, 0.00207437, 0.0020957 ,\n",
       "         0.00391857, 0.00118856, 0.00230416, 0.00281114, 0.00442782,\n",
       "         0.0015881 , 0.00745826, 0.0015965 , 0.0021638 , 0.00214091,\n",
       "         0.00161764, 0.00514841, 0.00279982, 0.00196707, 0.00524262,\n",
       "         0.00158327, 0.00100088, 0.00202392, 0.00387418, 0.00152139,\n",
       "         0.00151206, 0.00716418, 0.00489936, 0.0019699 , 0.00336328],\n",
       "        [0.00218368, 0.00209763, 0.00441022, 0.00222084, 0.00211042,\n",
       "         0.001714  , 0.00165781, 0.00979603, 0.00334506, 0.0056306 ,\n",
       "         0.00433817, 0.00260242, 0.0021696 , 0.00360582, 0.00221111,\n",
       "         0.00174567, 0.00201513, 0.00176302, 0.00242707, 0.00182754,\n",
       "         0.00273384, 0.00439774, 0.00220329, 0.00359295, 0.00474804,\n",
       "         0.00147336, 0.00929792, 0.00143062, 0.00572066, 0.00224921],\n",
       "        [0.00566491, 0.00661019, 0.0054153 , 0.00215879, 0.00176761,\n",
       "         0.00571894, 0.00469357, 0.0036881 , 0.00326901, 0.00523191,\n",
       "         0.00190031, 0.00750817, 0.00314911, 0.00232369, 0.00144236,\n",
       "         0.00209216, 0.00211145, 0.00424654, 0.00447402, 0.00329003,\n",
       "         0.00203624, 0.00307613, 0.0041928 , 0.00394207, 0.00836946,\n",
       "         0.00179823, 0.00496766, 0.00457853, 0.00279429, 0.00168206],\n",
       "        [0.00687583, 0.00147088, 0.00666172, 0.00400995, 0.00519511,\n",
       "         0.0019539 , 0.00226857, 0.00211126, 0.00292674, 0.00222733,\n",
       "         0.0020272 , 0.00416601, 0.0038624 , 0.00223911, 0.00400104,\n",
       "         0.00192922, 0.0022951 , 0.00689524, 0.00284972, 0.00430824,\n",
       "         0.00193489, 0.00391366, 0.00642208, 0.00718445, 0.00153731,\n",
       "         0.01052608, 0.0019937 , 0.00192113, 0.00621613, 0.00137313],\n",
       "        [0.00236742, 0.0021714 , 0.00133542, 0.00227059, 0.00217337,\n",
       "         0.00664698, 0.00876401, 0.00232659, 0.00504048, 0.00407829,\n",
       "         0.00324705, 0.00403846, 0.00566776, 0.00845151, 0.00742011,\n",
       "         0.00648395, 0.00294321, 0.00462923, 0.00176773, 0.00153202,\n",
       "         0.00228893, 0.00198374, 0.00234062, 0.0018417 , 0.00632038,\n",
       "         0.00118078, 0.0017166 , 0.00607118, 0.00287131, 0.00236422]],\n",
       "       dtype=float32))"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmf.train(link_X = 'sigmoid', link_Y = 'sigmoid', \n",
    "          weight_X = None, weight_Y =wY, \n",
    "          optim_steps=501, verbose=50, lr=0.05)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "narrow-logging",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}