RSNMF_test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = np.array([[5,0,4,0,3,3],\n",
    "            [0,4,0,0,3,0],\n",
    "            [3,0,0,3,0,0],\n",
    "            [0,4,0,5,0,1]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[5, 0, 4, 0, 3, 3],\n",
       "        [0, 4, 0, 0, 3, 0],\n",
       "        [3, 0, 0, 3, 0, 0],\n",
       "        [0, 4, 0, 5, 0, 1]])"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "R = np.asmatrix(x)\n",
    "R"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "U,I = R.shape\n",
    "f = 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#p = np.random.rand(U,f)\n",
    "#q = np.random.rand(f,I)\n",
    "p = np.array([[0.11,0.07,0.19],\n",
    "            [0.09,0.16,0.19],\n",
    "            [0.09,0.05,0.04],\n",
    "            [0.03,0.13,0.18]])\n",
    "q = np.array([[0.16,0.01,0.07,0.17,0.02,0.20],\n",
    "             [0.18,0.19,0.10,0.05,0.18,0.15],\n",
    "             [0.02,0.18,0.03,0.14,0.17,0.06]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "P = np.asmatrix(p)\n",
    "Q = np.asmatrix(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[0.11, 0.07, 0.19],\n",
       "        [0.09, 0.16, 0.19],\n",
       "        [0.09, 0.05, 0.04],\n",
       "        [0.03, 0.13, 0.18]])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "P"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[0.16, 0.01, 0.07, 0.17, 0.02, 0.2 ],\n",
       "        [0.18, 0.19, 0.1 , 0.05, 0.18, 0.15],\n",
       "        [0.02, 0.18, 0.03, 0.14, 0.17, 0.06]])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Q"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "lambda_p = 0.06\n",
    "lambda_q = 0.06"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "t = 1\n",
    "max_training_round = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "Ui = [2,2,1,2,2,2]\n",
    "Iu = [4,2,2,3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "UserUp = np.zeros((U,f), dtype=float)\n",
    "UserDown = np.zeros((U,f), dtype=float)\n",
    "ItemUp = np.zeros((f,I), dtype=float)\n",
    "ItemDown = np.zeros((f,I), dtype=float)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    " for u in range(U):\n",
    "        for i in range(I):\n",
    "            if(R[u,i] != 0):\n",
    "                rui_cap = P[u,:]*Q[:,i]\n",
    "                for k in range(f):\n",
    "                    UserUp[u,k] = UserUp[u,k] + Q[k,i] * R[u,i]\n",
    "                    UserDown[u,k] = UserDown[u,k] + Q[k,i] * rui_cap\n",
    "                    ItemUp[k,i] = ItemUp[k,i] + P[u,k] * R[u,i]\n",
    "                    ItemDown[k,i] = ItemDown[k,i] + P[u,k]*rui_cap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.74, 2.29, 0.91],\n",
       "       [0.1 , 1.3 , 1.23],\n",
       "       [0.99, 0.69, 0.48],\n",
       "       [1.09, 1.16, 1.48]])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "UserUp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.01659 , 0.023223, 0.011933],\n",
       "       [0.001913, 0.023767, 0.022483],\n",
       "       [0.00785 , 0.005526, 0.00376 ],\n",
       "       [0.01409 , 0.018191, 0.017662]])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "UserDown"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.82, 0.48, 0.44, 0.42, 0.6 , 0.36],\n",
       "       [0.5 , 1.16, 0.28, 0.8 , 0.69, 0.34],\n",
       "       [1.07, 1.48, 0.76, 1.02, 1.14, 0.75]])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ItemUp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.005918, 0.007617, 0.002244, 0.00321 , 0.010842, 0.005918],\n",
       "       [0.00359 , 0.017942, 0.001428, 0.005954, 0.013361, 0.007792],\n",
       "       [0.007428, 0.022777, 0.003876, 0.00756 , 0.0209  , 0.014875]])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ItemDown"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "for u in range(U):\n",
    "        for k in range(f):\n",
    "            UserDown[u,k] = UserDown[u,k] + Iu[u] * lambda_p * P[u,k]\n",
    "            P[u,k] = P[u,k] * (UserUp[u,k]/UserDown[u,k])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.04299 , 0.040023, 0.057533],\n",
       "       [0.012713, 0.042967, 0.045283],\n",
       "       [0.01865 , 0.011526, 0.00856 ],\n",
       "       [0.01949 , 0.041591, 0.050062]])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "UserDown"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[4.45219819, 4.00519701, 3.00523178],\n",
       "        [0.70793676, 4.84092443, 5.16087715],\n",
       "        [4.77747989, 2.99323269, 2.24299065],\n",
       "        [1.67778348, 3.62578442, 5.32140146]])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "P"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(I):\n",
    "        for k in range(f):\n",
    "            ItemDown[k,i] = ItemDown[k,i] + Ui[i] * lambda_q * Q[k,i]\n",
    "            Q[k,i] = Q[k,i] * (ItemUp[k,i]/ItemDown[k,i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.025118, 0.008817, 0.006444, 0.02361 , 0.013242, 0.029918],\n",
       "       [0.02519 , 0.040742, 0.007428, 0.011954, 0.034961, 0.025792],\n",
       "       [0.009828, 0.044377, 0.005676, 0.02436 , 0.0413  , 0.022075]])"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ItemDown"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[5.22334581, 0.54440286, 4.77963998, 3.02414231, 0.90620752,\n",
       "         2.40657798],\n",
       "        [3.57284637, 5.40965097, 3.76952073, 3.34616028, 3.55252996,\n",
       "         1.97735732],\n",
       "        [2.17745218, 6.00310972, 4.01691332, 5.86206897, 4.69249395,\n",
       "         2.0385051 ]])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Q"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}