docs/recommend/introduction.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Recommendation Systems\n",
    "\n",
    "There are many ways to recommend items to users. There are two primary types of recommendation systems, each with different sub-types. The two primary types are **content-based** and **collaborative filtering**.\n",
    "\n",
    "\n",
    "## Collaborative Filtering\n",
    "\n",
    "It primarily makes recommendations based on inputs or actions from other people.\n",
    "\n",
    "```{image} https://miro.medium.com/v2/resize:fit:1400/format:jpg/1*DqKuqlvMREPH18ccbs5hmg.png\n",
    ":alt: Collaborative Filtering\n",
    ":width: 80%\n",
    ":align: center\n",
    "```\n",
    "\n",
    "- Ignore User and Item Attributes\n",
    "- Focus on User-Item Interactions\n",
    "- Pure Behavior-Based Recommendation\n",
    "\n",
    "Variations on this type of recommendation system include:\n",
    "\n",
    "**Key Concepts**\n",
    "\n",
    "- Nearest Neighbor Collaborative Filtering\n",
    "- User-User CF Algorithm\n",
    "    - Neighborhoods and Tuning Parameters\n",
    "    - Alternatives to Historic Agreement (social, trust)\n",
    "- Item-Item CF Algorithm\n",
    "    - Dealing with Unary Data\n",
    "    - Hybrids and Extensions\n",
    "    - Practical Implications\n",
    "\n",
    "### User-User Collaborative Filtering\n",
    "\n",
    "This strategy involves creating user groups by comparing users’ activities and providing recommendations that are popular among other members of the group. It is useful on sites with a strong but versatile audience to quickly provide recommendations for a user on which little information is available.\n",
    "\n",
    "Find users similar to you and recommend what they like.\n",
    "\n",
    "#### Excerise: Movie Recommendations\n",
    "\n",
    "This is a 25 user x 100 movie matrix of ratings selected from the class data set. \n",
    "Rows are movies ratings, columns are users, and cells are ratings from 1 to 5.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import torch\n",
    "import seaborn\n",
    "import numpy as np\n",
    "from sklearn.metrics.pairwise import cosine_similarity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('https://github.com/akkefa/ml-notes/releases/download/v0.1.0/recommendation_systems_movies_ratings_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>1648</th>\n",
       "      <th>5136</th>\n",
       "      <th>918</th>\n",
       "      <th>2824</th>\n",
       "      <th>3867</th>\n",
       "      <th>860</th>\n",
       "      <th>3712</th>\n",
       "      <th>2968</th>\n",
       "      <th>3525</th>\n",
       "      <th>...</th>\n",
       "      <th>3556</th>\n",
       "      <th>5261</th>\n",
       "      <th>2492</th>\n",
       "      <th>5062</th>\n",
       "      <th>2486</th>\n",
       "      <th>4942</th>\n",
       "      <th>2267</th>\n",
       "      <th>4809</th>\n",
       "      <th>3853</th>\n",
       "      <th>2288</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>11: Star Wars: Episode IV - A New Hope (1977)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12: Finding Nemo (2003)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>13: Forrest Gump (1994)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>3.5</td>\n",
       "      <td>3.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>14: American Beauty (1999)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>22: Pirates of the Caribbean: The Curse of the...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          Unnamed: 0  1648  5136  918  2824   \n",
       "0      11: Star Wars: Episode IV - A New Hope (1977)   NaN   4.5  5.0   4.5  \\\n",
       "1                            12: Finding Nemo (2003)   NaN   5.0  5.0   NaN   \n",
       "2                            13: Forrest Gump (1994)   NaN   5.0  4.5   5.0   \n",
       "3                         14: American Beauty (1999)   NaN   4.0  NaN   NaN   \n",
       "4  22: Pirates of the Caribbean: The Curse of the...   4.0   5.0  3.0   4.5   \n",
       "\n",
       "   3867  860  3712  2968  3525  ...  3556  5261  2492  5062  2486  4942  2267   \n",
       "0   4.0  4.0   NaN   5.0   4.0  ...   4.0   NaN   4.5   4.0   3.5   NaN   NaN  \\\n",
       "1   4.0  4.0   4.5   4.5   4.0  ...   4.0   NaN   3.5   4.0   2.0   3.5   NaN   \n",
       "2   4.5  4.5   NaN   5.0   4.5  ...   4.0   5.0   3.5   4.5   4.5   4.0   3.5   \n",
       "3   NaN  NaN   4.5   2.0   3.5  ...   4.0   NaN   3.5   4.5   3.5   4.0   NaN   \n",
       "4   4.0  2.5   NaN   5.0   3.0  ...   3.0   1.5   4.0   4.0   2.5   3.5   NaN   \n",
       "\n",
       "   4809  3853  2288  \n",
       "0   NaN   NaN   NaN  \n",
       "1   NaN   NaN   3.5  \n",
       "2   4.5   3.5   3.5  \n",
       "3   3.5   NaN   NaN  \n",
       "4   5.0   NaN   3.5  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "tmp_df = df.copy()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop the first column (movie title)\n",
    "tmp_df.drop(columns=tmp_df.columns[0], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>1648</th>\n",
       "      <th>5136</th>\n",
       "      <th>918</th>\n",
       "      <th>2824</th>\n",
       "      <th>3867</th>\n",
       "      <th>860</th>\n",
       "      <th>3712</th>\n",
       "      <th>2968</th>\n",
       "      <th>3525</th>\n",
       "      <th>4323</th>\n",
       "      <th>...</th>\n",
       "      <th>3556</th>\n",
       "      <th>5261</th>\n",
       "      <th>2492</th>\n",
       "      <th>5062</th>\n",
       "      <th>2486</th>\n",
       "      <th>4942</th>\n",
       "      <th>2267</th>\n",
       "      <th>4809</th>\n",
       "      <th>3853</th>\n",
       "      <th>2288</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4.5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>3.5</td>\n",
       "      <td>3.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>3.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   1648  5136  918  2824  3867  860  3712  2968  3525  4323  ...  3556  5261   \n",
       "0   NaN   4.5  5.0   4.5   4.0  4.0   NaN   5.0   4.0   5.0  ...   4.0   NaN  \\\n",
       "1   NaN   5.0  5.0   NaN   4.0  4.0   4.5   4.5   4.0   5.0  ...   4.0   NaN   \n",
       "2   NaN   5.0  4.5   5.0   4.5  4.5   NaN   5.0   4.5   5.0  ...   4.0   5.0   \n",
       "3   NaN   4.0  NaN   NaN   NaN  NaN   4.5   2.0   3.5   5.0  ...   4.0   NaN   \n",
       "4   4.0   5.0  3.0   4.5   4.0  2.5   NaN   5.0   3.0   4.0  ...   3.0   1.5   \n",
       "\n",
       "   2492  5062  2486  4942  2267  4809  3853  2288  \n",
       "0   4.5   4.0   3.5   NaN   NaN   NaN   NaN   NaN  \n",
       "1   3.5   4.0   2.0   3.5   NaN   NaN   NaN   3.5  \n",
       "2   3.5   4.5   4.5   4.0   3.5   4.5   3.5   3.5  \n",
       "3   3.5   4.5   3.5   4.0   NaN   3.5   NaN   NaN  \n",
       "4   4.0   4.0   2.5   3.5   NaN   5.0   NaN   3.5  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tmp_df.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Given a set of items $I$, and a set of users $U$, and a sparse matrix of ratings $R$,\n",
    "We compute the prediction $s(\\mathrm{u}, \\mathrm{i})$ as follows:\n",
    "- For all users $v \\neq u$, compute $w_{u v}$\n",
    "- similarity metric (e.g., Pearson correlation)\n",
    "- Select a neighborhood of users $V \\subset U$ with highest $w_{u v}$\n",
    "- may limit neighborhood to top-k neighbors\n",
    "- may limit neighborhood to sim > sim_threshold\n",
    "- may use sim or |sim| (risks of negative correlations)\n",
    "- may limit neighborhood to people who rated i (if single-use)\n",
    "\n",
    "$$\n",
    "s(u, i)=\\bar{r}_u+\\frac{\\sum_{v \\in V}\\left(r_{v i}-\\bar{r}_v\\right) * w_{u v}}{\\sum_{v \\in V} w_{u v}}\n",
    "$$\n",
    "\n",
    "Computing the person correlation coefficient between each pair of users.\n",
    "Pearson correlation coefficient formula:\n",
    "\n",
    "$$\n",
    "r_{xy} = \\frac{\\sum_{i=1}^{n}(x_i - \\bar{x})(y_i - \\bar{y})}{\\sqrt{\\sum_{i=1}^{n}(x_i - \\bar{x})^2}\\sqrt{\\sum_{i=1}^{n}(y_i - \\bar{y})^2}}\n",
    "$$\n",
    "\n",
    "where $\\bar{x}$ and $\\bar{y}$ are the means of $x$ and $y$ respectively.\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>1648</th>\n",
       "      <th>5136</th>\n",
       "      <th>918</th>\n",
       "      <th>2824</th>\n",
       "      <th>3867</th>\n",
       "      <th>860</th>\n",
       "      <th>3712</th>\n",
       "      <th>2968</th>\n",
       "      <th>3525</th>\n",
       "      <th>4323</th>\n",
       "      <th>...</th>\n",
       "      <th>3556</th>\n",
       "      <th>5261</th>\n",
       "      <th>2492</th>\n",
       "      <th>5062</th>\n",
       "      <th>2486</th>\n",
       "      <th>4942</th>\n",
       "      <th>2267</th>\n",
       "      <th>4809</th>\n",
       "      <th>3853</th>\n",
       "      <th>2288</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1648</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.402980</td>\n",
       "      <td>-0.142206</td>\n",
       "      <td>0.517620</td>\n",
       "      <td>0.300200</td>\n",
       "      <td>0.480537</td>\n",
       "      <td>-0.312412</td>\n",
       "      <td>0.383348</td>\n",
       "      <td>0.092775</td>\n",
       "      <td>0.098191</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.191988</td>\n",
       "      <td>0.493008</td>\n",
       "      <td>0.360644</td>\n",
       "      <td>0.551089</td>\n",
       "      <td>0.002544</td>\n",
       "      <td>0.116653</td>\n",
       "      <td>-0.429183</td>\n",
       "      <td>0.394371</td>\n",
       "      <td>-0.304422</td>\n",
       "      <td>0.245048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5136</th>\n",
       "      <td>0.402980</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.118979</td>\n",
       "      <td>0.057916</td>\n",
       "      <td>0.341734</td>\n",
       "      <td>0.241377</td>\n",
       "      <td>0.131398</td>\n",
       "      <td>0.206695</td>\n",
       "      <td>0.360056</td>\n",
       "      <td>0.033642</td>\n",
       "      <td>...</td>\n",
       "      <td>0.488607</td>\n",
       "      <td>0.328120</td>\n",
       "      <td>0.422236</td>\n",
       "      <td>0.226635</td>\n",
       "      <td>0.305803</td>\n",
       "      <td>0.037769</td>\n",
       "      <td>0.240728</td>\n",
       "      <td>0.411676</td>\n",
       "      <td>0.189234</td>\n",
       "      <td>0.390067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>918</th>\n",
       "      <td>-0.142206</td>\n",
       "      <td>0.118979</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.317063</td>\n",
       "      <td>0.294558</td>\n",
       "      <td>0.468333</td>\n",
       "      <td>0.092037</td>\n",
       "      <td>-0.045854</td>\n",
       "      <td>0.367568</td>\n",
       "      <td>-0.035394</td>\n",
       "      <td>...</td>\n",
       "      <td>0.373226</td>\n",
       "      <td>0.470972</td>\n",
       "      <td>0.069956</td>\n",
       "      <td>-0.054762</td>\n",
       "      <td>0.133812</td>\n",
       "      <td>0.015169</td>\n",
       "      <td>-0.273096</td>\n",
       "      <td>0.082528</td>\n",
       "      <td>0.667168</td>\n",
       "      <td>0.119162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2824</th>\n",
       "      <td>0.517620</td>\n",
       "      <td>0.057916</td>\n",
       "      <td>-0.317063</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.060913</td>\n",
       "      <td>-0.008066</td>\n",
       "      <td>0.462910</td>\n",
       "      <td>0.214760</td>\n",
       "      <td>0.169907</td>\n",
       "      <td>0.119350</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.201275</td>\n",
       "      <td>0.228341</td>\n",
       "      <td>0.238700</td>\n",
       "      <td>0.259660</td>\n",
       "      <td>0.247097</td>\n",
       "      <td>0.149247</td>\n",
       "      <td>-0.361466</td>\n",
       "      <td>0.474974</td>\n",
       "      <td>-0.262073</td>\n",
       "      <td>0.166999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3867</th>\n",
       "      <td>0.300200</td>\n",
       "      <td>0.341734</td>\n",
       "      <td>0.294558</td>\n",
       "      <td>-0.060913</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.282497</td>\n",
       "      <td>0.400275</td>\n",
       "      <td>0.264249</td>\n",
       "      <td>0.125193</td>\n",
       "      <td>-0.333602</td>\n",
       "      <td>...</td>\n",
       "      <td>0.174085</td>\n",
       "      <td>0.297977</td>\n",
       "      <td>0.476683</td>\n",
       "      <td>0.293868</td>\n",
       "      <td>0.438992</td>\n",
       "      <td>-0.162818</td>\n",
       "      <td>-0.295966</td>\n",
       "      <td>0.054518</td>\n",
       "      <td>0.464110</td>\n",
       "      <td>0.379856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>860</th>\n",
       "      <td>0.480537</td>\n",
       "      <td>0.241377</td>\n",
       "      <td>0.468333</td>\n",
       "      <td>-0.008066</td>\n",
       "      <td>0.282497</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.171151</td>\n",
       "      <td>0.072927</td>\n",
       "      <td>0.387133</td>\n",
       "      <td>0.146158</td>\n",
       "      <td>...</td>\n",
       "      <td>0.347470</td>\n",
       "      <td>0.399436</td>\n",
       "      <td>0.207314</td>\n",
       "      <td>0.311363</td>\n",
       "      <td>0.276306</td>\n",
       "      <td>0.079698</td>\n",
       "      <td>0.212991</td>\n",
       "      <td>0.165608</td>\n",
       "      <td>0.162314</td>\n",
       "      <td>0.279677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3712</th>\n",
       "      <td>-0.312412</td>\n",
       "      <td>0.131398</td>\n",
       "      <td>0.092037</td>\n",
       "      <td>0.462910</td>\n",
       "      <td>0.400275</td>\n",
       "      <td>0.171151</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.065015</td>\n",
       "      <td>0.095623</td>\n",
       "      <td>-0.292501</td>\n",
       "      <td>...</td>\n",
       "      <td>0.016406</td>\n",
       "      <td>-0.240764</td>\n",
       "      <td>-0.115254</td>\n",
       "      <td>0.247693</td>\n",
       "      <td>0.166913</td>\n",
       "      <td>0.146011</td>\n",
       "      <td>0.009685</td>\n",
       "      <td>-0.451625</td>\n",
       "      <td>0.193660</td>\n",
       "      <td>0.113266</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2968</th>\n",
       "      <td>0.383348</td>\n",
       "      <td>0.206695</td>\n",
       "      <td>-0.045854</td>\n",
       "      <td>0.214760</td>\n",
       "      <td>0.264249</td>\n",
       "      <td>0.072927</td>\n",
       "      <td>0.065015</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.028529</td>\n",
       "      <td>-0.073252</td>\n",
       "      <td>...</td>\n",
       "      <td>0.049132</td>\n",
       "      <td>-0.009041</td>\n",
       "      <td>0.203613</td>\n",
       "      <td>0.033301</td>\n",
       "      <td>0.137982</td>\n",
       "      <td>0.070602</td>\n",
       "      <td>0.109452</td>\n",
       "      <td>-0.083562</td>\n",
       "      <td>-0.089317</td>\n",
       "      <td>0.229219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3525</th>\n",
       "      <td>0.092775</td>\n",
       "      <td>0.360056</td>\n",
       "      <td>0.367568</td>\n",
       "      <td>0.169907</td>\n",
       "      <td>0.125193</td>\n",
       "      <td>0.387133</td>\n",
       "      <td>0.095623</td>\n",
       "      <td>0.028529</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.210879</td>\n",
       "      <td>...</td>\n",
       "      <td>0.475711</td>\n",
       "      <td>0.306957</td>\n",
       "      <td>0.136343</td>\n",
       "      <td>0.301750</td>\n",
       "      <td>0.143414</td>\n",
       "      <td>0.056100</td>\n",
       "      <td>0.179908</td>\n",
       "      <td>0.284648</td>\n",
       "      <td>0.170757</td>\n",
       "      <td>0.193131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4323</th>\n",
       "      <td>0.098191</td>\n",
       "      <td>0.033642</td>\n",
       "      <td>-0.035394</td>\n",
       "      <td>0.119350</td>\n",
       "      <td>-0.333602</td>\n",
       "      <td>0.146158</td>\n",
       "      <td>-0.292501</td>\n",
       "      <td>-0.073252</td>\n",
       "      <td>0.210879</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.040606</td>\n",
       "      <td>0.155045</td>\n",
       "      <td>-0.204164</td>\n",
       "      <td>0.263654</td>\n",
       "      <td>0.167198</td>\n",
       "      <td>-0.084592</td>\n",
       "      <td>0.315712</td>\n",
       "      <td>0.085673</td>\n",
       "      <td>-0.109892</td>\n",
       "      <td>-0.279385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3617</th>\n",
       "      <td>-0.041734</td>\n",
       "      <td>0.138548</td>\n",
       "      <td>0.011316</td>\n",
       "      <td>0.282756</td>\n",
       "      <td>-0.066576</td>\n",
       "      <td>0.219929</td>\n",
       "      <td>-0.038900</td>\n",
       "      <td>0.312573</td>\n",
       "      <td>0.243283</td>\n",
       "      <td>0.022907</td>\n",
       "      <td>...</td>\n",
       "      <td>0.079571</td>\n",
       "      <td>-0.165628</td>\n",
       "      <td>0.053306</td>\n",
       "      <td>0.007810</td>\n",
       "      <td>-0.244637</td>\n",
       "      <td>-0.030709</td>\n",
       "      <td>-0.070660</td>\n",
       "      <td>0.268595</td>\n",
       "      <td>-0.143503</td>\n",
       "      <td>0.013284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4360</th>\n",
       "      <td>0.264425</td>\n",
       "      <td>0.152948</td>\n",
       "      <td>-0.231660</td>\n",
       "      <td>-0.005326</td>\n",
       "      <td>-0.093801</td>\n",
       "      <td>-0.005316</td>\n",
       "      <td>-0.364324</td>\n",
       "      <td>0.053024</td>\n",
       "      <td>-0.086061</td>\n",
       "      <td>0.252529</td>\n",
       "      <td>...</td>\n",
       "      <td>0.072993</td>\n",
       "      <td>0.161882</td>\n",
       "      <td>-0.000311</td>\n",
       "      <td>-0.077598</td>\n",
       "      <td>0.039389</td>\n",
       "      <td>-0.156091</td>\n",
       "      <td>0.408592</td>\n",
       "      <td>0.179652</td>\n",
       "      <td>0.280402</td>\n",
       "      <td>0.040328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2756</th>\n",
       "      <td>0.261268</td>\n",
       "      <td>0.148882</td>\n",
       "      <td>0.148431</td>\n",
       "      <td>-0.087747</td>\n",
       "      <td>0.310104</td>\n",
       "      <td>0.323499</td>\n",
       "      <td>0.126899</td>\n",
       "      <td>0.143347</td>\n",
       "      <td>0.058365</td>\n",
       "      <td>-0.221789</td>\n",
       "      <td>...</td>\n",
       "      <td>0.101784</td>\n",
       "      <td>-0.140953</td>\n",
       "      <td>0.150476</td>\n",
       "      <td>0.024572</td>\n",
       "      <td>-0.031130</td>\n",
       "      <td>-0.133768</td>\n",
       "      <td>0.142067</td>\n",
       "      <td>0.015140</td>\n",
       "      <td>0.181210</td>\n",
       "      <td>-0.005935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>0.464610</td>\n",
       "      <td>0.562449</td>\n",
       "      <td>0.267029</td>\n",
       "      <td>0.241567</td>\n",
       "      <td>-0.003878</td>\n",
       "      <td>0.539066</td>\n",
       "      <td>-0.051320</td>\n",
       "      <td>-0.118085</td>\n",
       "      <td>0.475495</td>\n",
       "      <td>0.258866</td>\n",
       "      <td>...</td>\n",
       "      <td>0.326774</td>\n",
       "      <td>0.291476</td>\n",
       "      <td>0.372676</td>\n",
       "      <td>0.525990</td>\n",
       "      <td>0.123380</td>\n",
       "      <td>0.178088</td>\n",
       "      <td>0.088600</td>\n",
       "      <td>0.668516</td>\n",
       "      <td>0.179680</td>\n",
       "      <td>0.155869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>442</th>\n",
       "      <td>0.022308</td>\n",
       "      <td>0.414438</td>\n",
       "      <td>0.304139</td>\n",
       "      <td>0.116532</td>\n",
       "      <td>0.113581</td>\n",
       "      <td>0.181276</td>\n",
       "      <td>0.227130</td>\n",
       "      <td>0.100841</td>\n",
       "      <td>0.201734</td>\n",
       "      <td>-0.024337</td>\n",
       "      <td>...</td>\n",
       "      <td>0.251660</td>\n",
       "      <td>0.046822</td>\n",
       "      <td>0.218575</td>\n",
       "      <td>0.150431</td>\n",
       "      <td>0.280392</td>\n",
       "      <td>0.038378</td>\n",
       "      <td>0.262520</td>\n",
       "      <td>0.064179</td>\n",
       "      <td>-0.023439</td>\n",
       "      <td>0.257864</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3556</th>\n",
       "      <td>-0.191988</td>\n",
       "      <td>0.488607</td>\n",
       "      <td>0.373226</td>\n",
       "      <td>-0.201275</td>\n",
       "      <td>0.174085</td>\n",
       "      <td>0.347470</td>\n",
       "      <td>0.016406</td>\n",
       "      <td>0.049132</td>\n",
       "      <td>0.475711</td>\n",
       "      <td>-0.040606</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.086665</td>\n",
       "      <td>0.158739</td>\n",
       "      <td>-0.016164</td>\n",
       "      <td>0.256537</td>\n",
       "      <td>-0.055137</td>\n",
       "      <td>0.503247</td>\n",
       "      <td>0.100277</td>\n",
       "      <td>0.423225</td>\n",
       "      <td>0.222458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5261</th>\n",
       "      <td>0.493008</td>\n",
       "      <td>0.328120</td>\n",
       "      <td>0.470972</td>\n",
       "      <td>0.228341</td>\n",
       "      <td>0.297977</td>\n",
       "      <td>0.399436</td>\n",
       "      <td>-0.240764</td>\n",
       "      <td>-0.009041</td>\n",
       "      <td>0.306957</td>\n",
       "      <td>0.155045</td>\n",
       "      <td>...</td>\n",
       "      <td>0.086665</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.149165</td>\n",
       "      <td>0.372177</td>\n",
       "      <td>0.198086</td>\n",
       "      <td>0.270928</td>\n",
       "      <td>-0.393376</td>\n",
       "      <td>0.455274</td>\n",
       "      <td>0.039050</td>\n",
       "      <td>0.374264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2492</th>\n",
       "      <td>0.360644</td>\n",
       "      <td>0.422236</td>\n",
       "      <td>0.069956</td>\n",
       "      <td>0.238700</td>\n",
       "      <td>0.476683</td>\n",
       "      <td>0.207314</td>\n",
       "      <td>-0.115254</td>\n",
       "      <td>0.203613</td>\n",
       "      <td>0.136343</td>\n",
       "      <td>-0.204164</td>\n",
       "      <td>...</td>\n",
       "      <td>0.158739</td>\n",
       "      <td>0.149165</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.276883</td>\n",
       "      <td>0.158002</td>\n",
       "      <td>0.035825</td>\n",
       "      <td>-0.345495</td>\n",
       "      <td>0.449025</td>\n",
       "      <td>0.289410</td>\n",
       "      <td>0.169239</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5062</th>\n",
       "      <td>0.551089</td>\n",
       "      <td>0.226635</td>\n",
       "      <td>-0.054762</td>\n",
       "      <td>0.259660</td>\n",
       "      <td>0.293868</td>\n",
       "      <td>0.311363</td>\n",
       "      <td>0.247693</td>\n",
       "      <td>0.033301</td>\n",
       "      <td>0.301750</td>\n",
       "      <td>0.263654</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.016164</td>\n",
       "      <td>0.372177</td>\n",
       "      <td>0.276883</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.403809</td>\n",
       "      <td>0.028521</td>\n",
       "      <td>0.107821</td>\n",
       "      <td>0.428055</td>\n",
       "      <td>0.407044</td>\n",
       "      <td>0.278868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2486</th>\n",
       "      <td>0.002544</td>\n",
       "      <td>0.305803</td>\n",
       "      <td>0.133812</td>\n",
       "      <td>0.247097</td>\n",
       "      <td>0.438992</td>\n",
       "      <td>0.276306</td>\n",
       "      <td>0.166913</td>\n",
       "      <td>0.137982</td>\n",
       "      <td>0.143414</td>\n",
       "      <td>0.167198</td>\n",
       "      <td>...</td>\n",
       "      <td>0.256537</td>\n",
       "      <td>0.198086</td>\n",
       "      <td>0.158002</td>\n",
       "      <td>0.403809</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.068421</td>\n",
       "      <td>0.173797</td>\n",
       "      <td>0.105761</td>\n",
       "      <td>0.472361</td>\n",
       "      <td>0.257462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4942</th>\n",
       "      <td>0.116653</td>\n",
       "      <td>0.037769</td>\n",
       "      <td>0.015169</td>\n",
       "      <td>0.149247</td>\n",
       "      <td>-0.162818</td>\n",
       "      <td>0.079698</td>\n",
       "      <td>0.146011</td>\n",
       "      <td>0.070602</td>\n",
       "      <td>0.056100</td>\n",
       "      <td>-0.084592</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.055137</td>\n",
       "      <td>0.270928</td>\n",
       "      <td>0.035825</td>\n",
       "      <td>0.028521</td>\n",
       "      <td>-0.068421</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.346386</td>\n",
       "      <td>-0.004638</td>\n",
       "      <td>0.143672</td>\n",
       "      <td>0.074476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2267</th>\n",
       "      <td>-0.429183</td>\n",
       "      <td>0.240728</td>\n",
       "      <td>-0.273096</td>\n",
       "      <td>-0.361466</td>\n",
       "      <td>-0.295966</td>\n",
       "      <td>0.212991</td>\n",
       "      <td>0.009685</td>\n",
       "      <td>0.109452</td>\n",
       "      <td>0.179908</td>\n",
       "      <td>0.315712</td>\n",
       "      <td>...</td>\n",
       "      <td>0.503247</td>\n",
       "      <td>-0.393376</td>\n",
       "      <td>-0.345495</td>\n",
       "      <td>0.107821</td>\n",
       "      <td>0.173797</td>\n",
       "      <td>-0.346386</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.339845</td>\n",
       "      <td>0.165960</td>\n",
       "      <td>0.156341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4809</th>\n",
       "      <td>0.394371</td>\n",
       "      <td>0.411676</td>\n",
       "      <td>0.082528</td>\n",
       "      <td>0.474974</td>\n",
       "      <td>0.054518</td>\n",
       "      <td>0.165608</td>\n",
       "      <td>-0.451625</td>\n",
       "      <td>-0.083562</td>\n",
       "      <td>0.284648</td>\n",
       "      <td>0.085673</td>\n",
       "      <td>...</td>\n",
       "      <td>0.100277</td>\n",
       "      <td>0.455274</td>\n",
       "      <td>0.449025</td>\n",
       "      <td>0.428055</td>\n",
       "      <td>0.105761</td>\n",
       "      <td>-0.004638</td>\n",
       "      <td>-0.339845</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.542192</td>\n",
       "      <td>0.435520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3853</th>\n",
       "      <td>-0.304422</td>\n",
       "      <td>0.189234</td>\n",
       "      <td>0.667168</td>\n",
       "      <td>-0.262073</td>\n",
       "      <td>0.464110</td>\n",
       "      <td>0.162314</td>\n",
       "      <td>0.193660</td>\n",
       "      <td>-0.089317</td>\n",
       "      <td>0.170757</td>\n",
       "      <td>-0.109892</td>\n",
       "      <td>...</td>\n",
       "      <td>0.423225</td>\n",
       "      <td>0.039050</td>\n",
       "      <td>0.289410</td>\n",
       "      <td>0.407044</td>\n",
       "      <td>0.472361</td>\n",
       "      <td>0.143672</td>\n",
       "      <td>0.165960</td>\n",
       "      <td>0.542192</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.080403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2288</th>\n",
       "      <td>0.245048</td>\n",
       "      <td>0.390067</td>\n",
       "      <td>0.119162</td>\n",
       "      <td>0.166999</td>\n",
       "      <td>0.379856</td>\n",
       "      <td>0.279677</td>\n",
       "      <td>0.113266</td>\n",
       "      <td>0.229219</td>\n",
       "      <td>0.193131</td>\n",
       "      <td>-0.279385</td>\n",
       "      <td>...</td>\n",
       "      <td>0.222458</td>\n",
       "      <td>0.374264</td>\n",
       "      <td>0.169239</td>\n",
       "      <td>0.278868</td>\n",
       "      <td>0.257462</td>\n",
       "      <td>0.074476</td>\n",
       "      <td>0.156341</td>\n",
       "      <td>0.435520</td>\n",
       "      <td>0.080403</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>25 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          1648      5136       918      2824      3867       860      3712   \n",
       "1648  1.000000  0.402980 -0.142206  0.517620  0.300200  0.480537 -0.312412  \\\n",
       "5136  0.402980  1.000000  0.118979  0.057916  0.341734  0.241377  0.131398   \n",
       "918  -0.142206  0.118979  1.000000 -0.317063  0.294558  0.468333  0.092037   \n",
       "2824  0.517620  0.057916 -0.317063  1.000000 -0.060913 -0.008066  0.462910   \n",
       "3867  0.300200  0.341734  0.294558 -0.060913  1.000000  0.282497  0.400275   \n",
       "860   0.480537  0.241377  0.468333 -0.008066  0.282497  1.000000  0.171151   \n",
       "3712 -0.312412  0.131398  0.092037  0.462910  0.400275  0.171151  1.000000   \n",
       "2968  0.383348  0.206695 -0.045854  0.214760  0.264249  0.072927  0.065015   \n",
       "3525  0.092775  0.360056  0.367568  0.169907  0.125193  0.387133  0.095623   \n",
       "4323  0.098191  0.033642 -0.035394  0.119350 -0.333602  0.146158 -0.292501   \n",
       "3617 -0.041734  0.138548  0.011316  0.282756 -0.066576  0.219929 -0.038900   \n",
       "4360  0.264425  0.152948 -0.231660 -0.005326 -0.093801 -0.005316 -0.364324   \n",
       "2756  0.261268  0.148882  0.148431 -0.087747  0.310104  0.323499  0.126899   \n",
       "89    0.464610  0.562449  0.267029  0.241567 -0.003878  0.539066 -0.051320   \n",
       "442   0.022308  0.414438  0.304139  0.116532  0.113581  0.181276  0.227130   \n",
       "3556 -0.191988  0.488607  0.373226 -0.201275  0.174085  0.347470  0.016406   \n",
       "5261  0.493008  0.328120  0.470972  0.228341  0.297977  0.399436 -0.240764   \n",
       "2492  0.360644  0.422236  0.069956  0.238700  0.476683  0.207314 -0.115254   \n",
       "5062  0.551089  0.226635 -0.054762  0.259660  0.293868  0.311363  0.247693   \n",
       "2486  0.002544  0.305803  0.133812  0.247097  0.438992  0.276306  0.166913   \n",
       "4942  0.116653  0.037769  0.015169  0.149247 -0.162818  0.079698  0.146011   \n",
       "2267 -0.429183  0.240728 -0.273096 -0.361466 -0.295966  0.212991  0.009685   \n",
       "4809  0.394371  0.411676  0.082528  0.474974  0.054518  0.165608 -0.451625   \n",
       "3853 -0.304422  0.189234  0.667168 -0.262073  0.464110  0.162314  0.193660   \n",
       "2288  0.245048  0.390067  0.119162  0.166999  0.379856  0.279677  0.113266   \n",
       "\n",
       "          2968      3525      4323  ...      3556      5261      2492   \n",
       "1648  0.383348  0.092775  0.098191  ... -0.191988  0.493008  0.360644  \\\n",
       "5136  0.206695  0.360056  0.033642  ...  0.488607  0.328120  0.422236   \n",
       "918  -0.045854  0.367568 -0.035394  ...  0.373226  0.470972  0.069956   \n",
       "2824  0.214760  0.169907  0.119350  ... -0.201275  0.228341  0.238700   \n",
       "3867  0.264249  0.125193 -0.333602  ...  0.174085  0.297977  0.476683   \n",
       "860   0.072927  0.387133  0.146158  ...  0.347470  0.399436  0.207314   \n",
       "3712  0.065015  0.095623 -0.292501  ...  0.016406 -0.240764 -0.115254   \n",
       "2968  1.000000  0.028529 -0.073252  ...  0.049132 -0.009041  0.203613   \n",
       "3525  0.028529  1.000000  0.210879  ...  0.475711  0.306957  0.136343   \n",
       "4323 -0.073252  0.210879  1.000000  ... -0.040606  0.155045 -0.204164   \n",
       "3617  0.312573  0.243283  0.022907  ...  0.079571 -0.165628  0.053306   \n",
       "4360  0.053024 -0.086061  0.252529  ...  0.072993  0.161882 -0.000311   \n",
       "2756  0.143347  0.058365 -0.221789  ...  0.101784 -0.140953  0.150476   \n",
       "89   -0.118085  0.475495  0.258866  ...  0.326774  0.291476  0.372676   \n",
       "442   0.100841  0.201734 -0.024337  ...  0.251660  0.046822  0.218575   \n",
       "3556  0.049132  0.475711 -0.040606  ...  1.000000  0.086665  0.158739   \n",
       "5261 -0.009041  0.306957  0.155045  ...  0.086665  1.000000  0.149165   \n",
       "2492  0.203613  0.136343 -0.204164  ...  0.158739  0.149165  1.000000   \n",
       "5062  0.033301  0.301750  0.263654  ... -0.016164  0.372177  0.276883   \n",
       "2486  0.137982  0.143414  0.167198  ...  0.256537  0.198086  0.158002   \n",
       "4942  0.070602  0.056100 -0.084592  ... -0.055137  0.270928  0.035825   \n",
       "2267  0.109452  0.179908  0.315712  ...  0.503247 -0.393376 -0.345495   \n",
       "4809 -0.083562  0.284648  0.085673  ...  0.100277  0.455274  0.449025   \n",
       "3853 -0.089317  0.170757 -0.109892  ...  0.423225  0.039050  0.289410   \n",
       "2288  0.229219  0.193131 -0.279385  ...  0.222458  0.374264  0.169239   \n",
       "\n",
       "          5062      2486      4942      2267      4809      3853      2288  \n",
       "1648  0.551089  0.002544  0.116653 -0.429183  0.394371 -0.304422  0.245048  \n",
       "5136  0.226635  0.305803  0.037769  0.240728  0.411676  0.189234  0.390067  \n",
       "918  -0.054762  0.133812  0.015169 -0.273096  0.082528  0.667168  0.119162  \n",
       "2824  0.259660  0.247097  0.149247 -0.361466  0.474974 -0.262073  0.166999  \n",
       "3867  0.293868  0.438992 -0.162818 -0.295966  0.054518  0.464110  0.379856  \n",
       "860   0.311363  0.276306  0.079698  0.212991  0.165608  0.162314  0.279677  \n",
       "3712  0.247693  0.166913  0.146011  0.009685 -0.451625  0.193660  0.113266  \n",
       "2968  0.033301  0.137982  0.070602  0.109452 -0.083562 -0.089317  0.229219  \n",
       "3525  0.301750  0.143414  0.056100  0.179908  0.284648  0.170757  0.193131  \n",
       "4323  0.263654  0.167198 -0.084592  0.315712  0.085673 -0.109892 -0.279385  \n",
       "3617  0.007810 -0.244637 -0.030709 -0.070660  0.268595 -0.143503  0.013284  \n",
       "4360 -0.077598  0.039389 -0.156091  0.408592  0.179652  0.280402  0.040328  \n",
       "2756  0.024572 -0.031130 -0.133768  0.142067  0.015140  0.181210 -0.005935  \n",
       "89    0.525990  0.123380  0.178088  0.088600  0.668516  0.179680  0.155869  \n",
       "442   0.150431  0.280392  0.038378  0.262520  0.064179 -0.023439  0.257864  \n",
       "3556 -0.016164  0.256537 -0.055137  0.503247  0.100277  0.423225  0.222458  \n",
       "5261  0.372177  0.198086  0.270928 -0.393376  0.455274  0.039050  0.374264  \n",
       "2492  0.276883  0.158002  0.035825 -0.345495  0.449025  0.289410  0.169239  \n",
       "5062  1.000000  0.403809  0.028521  0.107821  0.428055  0.407044  0.278868  \n",
       "2486  0.403809  1.000000 -0.068421  0.173797  0.105761  0.472361  0.257462  \n",
       "4942  0.028521 -0.068421  1.000000 -0.346386 -0.004638  0.143672  0.074476  \n",
       "2267  0.107821  0.173797 -0.346386  1.000000 -0.339845  0.165960  0.156341  \n",
       "4809  0.428055  0.105761 -0.004638 -0.339845  1.000000  0.542192  0.435520  \n",
       "3853  0.407044  0.472361  0.143672  0.165960  0.542192  1.000000  0.080403  \n",
       "2288  0.278868  0.257462  0.074476  0.156341  0.435520  0.080403  1.000000  \n",
       "\n",
       "[25 rows x 25 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_df = tmp_df.corr()\n",
    "corr_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "seaborn.heatmap(corr_df.corr())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3867    1.000000\n",
       "2492    0.476683\n",
       "3853    0.464110\n",
       "2486    0.438992\n",
       "3712    0.400275\n",
       "2288    0.379856\n",
       "5136    0.341734\n",
       "2756    0.310104\n",
       "1648    0.300200\n",
       "5261    0.297977\n",
       "918     0.294558\n",
       "5062    0.293868\n",
       "860     0.282497\n",
       "2968    0.264249\n",
       "3556    0.174085\n",
       "3525    0.125193\n",
       "442     0.113581\n",
       "4809    0.054518\n",
       "89     -0.003878\n",
       "2824   -0.060913\n",
       "3617   -0.066576\n",
       "4360   -0.093801\n",
       "4942   -0.162818\n",
       "2267   -0.295966\n",
       "4323   -0.333602\n",
       "Name: 3867, dtype: float64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_df[\"3867\"].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "89      1.000000\n",
       "4809    0.668516\n",
       "5136    0.562449\n",
       "860     0.539066\n",
       "5062    0.525990\n",
       "3525    0.475495\n",
       "1648    0.464610\n",
       "2492    0.372676\n",
       "3556    0.326774\n",
       "442     0.296826\n",
       "5261    0.291476\n",
       "2756    0.290591\n",
       "3617    0.278335\n",
       "918     0.267029\n",
       "4323    0.258866\n",
       "2824    0.241567\n",
       "3853    0.179680\n",
       "4942    0.178088\n",
       "2288    0.155869\n",
       "2486    0.123380\n",
       "2267    0.088600\n",
       "3867   -0.003878\n",
       "3712   -0.051320\n",
       "4360   -0.115492\n",
       "2968   -0.118085\n",
       "Name: 89, dtype: float64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_df[\"89\"].sort_values(ascending=False)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compute the predictions for each movie for users 3867 and 89 by taking the correlation-weighted average of the ratings of the top-five neighbors (for each target user) for each movie. The formal formula for correlation-weighted average is\n",
    "\n",
    "$$\n",
    "\\hat{x}_{u,i} = \\frac{\\sum_{v \\in N}r_{u,v}x_{v,i}}{\\sum_{v \\in N}|r_{u,v}|}\n",
    "$$\n",
    "\n",
    "where $N$ is the set of the top-five neighbors of user $u$ and $x_{v,i}$ is the rating of user $v$ for movie $i$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_top_users(df_corr,target,n=5):\n",
    "    target_cor = df_corr.loc[target]\n",
    "    top_neighbors = target_cor.nlargest(n+1).iloc[1:]\n",
    "    return top_neighbors\n",
    "\n",
    "def get_user_movie_score(movie,user):\n",
    "    neighbors = get_top_users(corr_df,str(user))\n",
    "    rating_sum = 0\n",
    "    weight_sum = 0\n",
    "    for user,w in zip(neighbors.index,neighbors.values):\n",
    "        if np.isnan(movie[user]):\n",
    "            continue\n",
    "        rating_sum += movie[user] * w\n",
    "        weight_sum += w\n",
    "    if weight_sum == 0:\n",
    "        return 0\n",
    "    else:\n",
    "        return rating_sum/weight_sum"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2492    0.476683\n",
       "3853    0.464110\n",
       "2486    0.438992\n",
       "3712    0.400275\n",
       "2288    0.379856\n",
       "Name: 3867, dtype: float64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_top_users(corr_df,\"3867\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2824    0.462910\n",
       "3867    0.400275\n",
       "5062    0.247693\n",
       "442     0.227130\n",
       "3853    0.193660\n",
       "Name: 3712, dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_top_users(corr_df,\"3712\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_3867 = df.apply(get_user_movie_score,axis=1,args=(3867,))\n",
    "pred_89 = df.apply(get_user_movie_score,axis=1,args=(89,))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "77    4.760291\n",
       "21    4.551454\n",
       "16    4.507637\n",
       "dtype: float64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_3867.sort_values(ascending=False)[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1891: Star Wars: Episode V - The Empire Strikes Back (1980)\n",
      "155: The Dark Knight (2008)\n",
      "122: The Lord of the Rings: The Return of the King (2003)\n",
      "77: Memento (2000)\n",
      "121: The Lord of the Rings: The Two Towers (2002)\n"
     ]
    }
   ],
   "source": [
    "for i in pred_3867.sort_values(ascending=False)[:5].index:\n",
    "    print(df.loc[i][0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "238: The Godfather (1972)\n",
      "278: The Shawshank Redemption (1994)\n",
      "807: Seven (a.k.a. Se7en) (1995)\n",
      "275: Fargo (1996)\n",
      "424: Schindler's List (1993)\n"
     ]
    }
   ],
   "source": [
    "for i in pred_89.sort_values(ascending=False)[:5].index:\n",
    "    print(df.loc[i][0])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Normalization\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_norm_user_movie_score(movie,user):\n",
    "    user = str(user)\n",
    "    neighbors = get_top_users(corr_df,str(user))\n",
    "    rating_sum = 0\n",
    "    weight_sum = 0\n",
    "    user_rating_mean = df.loc[:,user].mean()\n",
    "    for user,w in zip(neighbors.index,neighbors.values):\n",
    "        if np.isnan(movie[user]):\n",
    "            continue\n",
    "        movie_user_mean = df.loc[:,user].mean()\n",
    "        rating_sum += (movie[user]-movie_user_mean) * w\n",
    "        weight_sum += w\n",
    "    if weight_sum == 0:\n",
    "        return 0\n",
    "    else:\n",
    "        return user_rating_mean + rating_sum/weight_sum  \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "norm_pred_3867 = df.apply(get_norm_user_movie_score,axis=1,args=(3867,))\n",
    "norm_pred_89 = df.apply(get_norm_user_movie_score,axis=1,args=(89,))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1891: Star Wars: Episode V - The Empire Strikes Back (1980)\n",
      "155: The Dark Knight (2008)\n",
      "77: Memento (2000)\n",
      "275: Fargo (1996)\n",
      "807: Seven (a.k.a. Se7en) (1995)\n"
     ]
    }
   ],
   "source": [
    "for i in norm_pred_3867.sort_values(ascending=False)[:5].index:\n",
    "    print(df.loc[i][0])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Problem in User - User Collaborative Filtering:\n",
    "\n",
    "Issues of Sparsity\n",
    "    – With large item sets, small numbers of ratings, too\n",
    "    often there are points where no recommendation\n",
    "    can be made (for a user, for an item to a set of\n",
    "    users, etc.)\n",
    "    – Many solutions proposed here, including\n",
    "    “filterbots”, item‐item, and dimensionality\n",
    "    reduction\n",
    "\n",
    "Computational performance\n",
    "    – With millions of users (or more), computing all‐\n",
    "    pairs correlations is expensive\n",
    "    – Even incremental approaches were expensive\n",
    "    – And user profiles could change quickly – needed\n",
    "    to compute in real time to keep users happy\n",
    "\n",
    "### Item-Item Collaborative Filtering\n",
    "\n",
    "Item‐Item similarity is fairly stable.\n",
    "\n",
    "- This is dependent on having many more usersthan items\n",
    "  - Average item has many more ratings than an average user\n",
    "  - Intuitively, items don’t generally change rapidly – at least not in ratings space (special case for time‐bound items)\n",
    "- Item similarity is a route to computing a prediction of a user’s item preference\n",
    "\n",
    "\n",
    "https://github.com/shenweichen/Coursera/blob/master/Specialization_Recommender_System_University_of_Minnesota/Course2_Nearest_Neighbor_Collaborative_Filtering/Item%20Based%20Assignment.ipynb\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_excel(\"https://github.com/akkefa/ml-notes/releases/download/v0.1.0/item_item_cb.xls\", sheet_name=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>User</th>\n",
       "      <th>1: Toy Story (1995)</th>\n",
       "      <th>1210: Star Wars: Episode VI - Return of the Jedi (1983)</th>\n",
       "      <th>356: Forrest Gump (1994)</th>\n",
       "      <th>318: Shawshank Redemption, The (1994)</th>\n",
       "      <th>593: Silence of the Lambs, The (1991)</th>\n",
       "      <th>3578: Gladiator (2000)</th>\n",
       "      <th>260: Star Wars: Episode IV - A New Hope (1977)</th>\n",
       "      <th>2028: Saving Private Ryan (1998)</th>\n",
       "      <th>296: Pulp Fiction (1994)</th>\n",
       "      <th>...</th>\n",
       "      <th>2916: Total Recall (1990)</th>\n",
       "      <th>780: Independence Day (ID4) (1996)</th>\n",
       "      <th>541: Blade Runner (1982)</th>\n",
       "      <th>1265: Groundhog Day (1993)</th>\n",
       "      <th>2571: Matrix, The (1999)</th>\n",
       "      <th>527: Schindler's List (1993)</th>\n",
       "      <th>2762: Sixth Sense, The (1999)</th>\n",
       "      <th>1198: Raiders of the Lost Ark (1981)</th>\n",
       "      <th>34: Babe (1995)</th>\n",
       "      <th>Mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>755</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5277</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.769231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1577</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4388</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.833333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1202</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.214286</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   User  1: Toy Story (1995)   \n",
       "0   755                  2.0  \\\n",
       "1  5277                  1.0   \n",
       "2  1577                  0.0   \n",
       "3  4388                  2.0   \n",
       "4  1202                  0.0   \n",
       "\n",
       "   1210: Star Wars: Episode VI - Return of the Jedi (1983)   \n",
       "0                                                5.0        \\\n",
       "1                                                0.0         \n",
       "2                                                0.0         \n",
       "3                                                3.0         \n",
       "4                                                3.0         \n",
       "\n",
       "   356: Forrest Gump (1994)  318: Shawshank Redemption, The (1994)   \n",
       "0                       2.0                                    0.0  \\\n",
       "1                       0.0                                    2.0   \n",
       "2                       0.0                                    5.0   \n",
       "3                       0.0                                    0.0   \n",
       "4                       4.0                                    1.0   \n",
       "\n",
       "   593: Silence of the Lambs, The (1991)  3578: Gladiator (2000)   \n",
       "0                                    4.0                     4.0  \\\n",
       "1                                    4.0                     2.0   \n",
       "2                                    2.0                     0.0   \n",
       "3                                    0.0                     1.0   \n",
       "4                                    4.0                     1.0   \n",
       "\n",
       "   260: Star Wars: Episode IV - A New Hope (1977)   \n",
       "0                                             1.0  \\\n",
       "1                                             5.0   \n",
       "2                                             0.0   \n",
       "3                                             0.0   \n",
       "4                                             4.0   \n",
       "\n",
       "   2028: Saving Private Ryan (1998)  296: Pulp Fiction (1994)  ...   \n",
       "0                               2.0                       0.0  ...  \\\n",
       "1                               0.0                       0.0  ...   \n",
       "2                               0.0                       0.0  ...   \n",
       "3                               3.0                       4.0  ...   \n",
       "4                               4.0                       0.0  ...   \n",
       "\n",
       "   2916: Total Recall (1990)  780: Independence Day (ID4) (1996)   \n",
       "0                        0.0                                 5.0  \\\n",
       "1                        2.0                                 2.0   \n",
       "2                        1.0                                 4.0   \n",
       "3                        4.0                                 0.0   \n",
       "4                        1.0                                 0.0   \n",
       "\n",
       "   541: Blade Runner (1982)  1265: Groundhog Day (1993)   \n",
       "0                       2.0                         5.0  \\\n",
       "1                       0.0                         2.0   \n",
       "2                       4.0                         1.0   \n",
       "3                       3.0                         5.0   \n",
       "4                       4.0                         0.0   \n",
       "\n",
       "   2571: Matrix, The (1999)  527: Schindler's List (1993)   \n",
       "0                       4.0                           2.0  \\\n",
       "1                       0.0                           5.0   \n",
       "2                       1.0                           2.0   \n",
       "3                       0.0                           5.0   \n",
       "4                       3.0                           5.0   \n",
       "\n",
       "   2762: Sixth Sense, The (1999)  1198: Raiders of the Lost Ark (1981)   \n",
       "0                            5.0                                   0.0  \\\n",
       "1                            1.0                                   3.0   \n",
       "2                            3.0                                   1.0   \n",
       "3                            1.0                                   1.0   \n",
       "4                            5.0                                   0.0   \n",
       "\n",
       "   34: Babe (1995)      Mean  \n",
       "0              0.0  3.200000  \n",
       "1              0.0  2.769231  \n",
       "2              3.0  2.333333  \n",
       "3              2.0  2.833333  \n",
       "4              0.0  3.214286  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "matrix = pd.read_excel('https://github.com/akkefa/ml-notes/releases/download/v0.1.0/item_item_cb.xls',sheet_name=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>1: Toy Story (1995)</th>\n",
       "      <th>1210: Star Wars: Episode VI - Return of the Jedi (1983)</th>\n",
       "      <th>356: Forrest Gump (1994)</th>\n",
       "      <th>318: Shawshank Redemption, The (1994)</th>\n",
       "      <th>593: Silence of the Lambs, The (1991)</th>\n",
       "      <th>3578: Gladiator (2000)</th>\n",
       "      <th>260: Star Wars: Episode IV - A New Hope (1977)</th>\n",
       "      <th>2028: Saving Private Ryan (1998)</th>\n",
       "      <th>296: Pulp Fiction (1994)</th>\n",
       "      <th>...</th>\n",
       "      <th>2396: Shakespeare in Love (1998)</th>\n",
       "      <th>2916: Total Recall (1990)</th>\n",
       "      <th>780: Independence Day (ID4) (1996)</th>\n",
       "      <th>541: Blade Runner (1982)</th>\n",
       "      <th>1265: Groundhog Day (1993)</th>\n",
       "      <th>2571: Matrix, The (1999)</th>\n",
       "      <th>527: Schindler's List (1993)</th>\n",
       "      <th>2762: Sixth Sense, The (1999)</th>\n",
       "      <th>1198: Raiders of the Lost Ark (1981)</th>\n",
       "      <th>34: Babe (1995)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1: Toy Story (1995)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1210: Star Wars: Episode VI - Return of the Je...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>356: Forrest Gump (1994)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>318: Shawshank Redemption, The (1994)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>593: Silence of the Lambs, The (1991)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          Unnamed: 0  1: Toy Story (1995)   \n",
       "0                                1: Toy Story (1995)                  NaN  \\\n",
       "1  1210: Star Wars: Episode VI - Return of the Je...                  NaN   \n",
       "2                           356: Forrest Gump (1994)                  NaN   \n",
       "3              318: Shawshank Redemption, The (1994)                  NaN   \n",
       "4              593: Silence of the Lambs, The (1991)                  NaN   \n",
       "\n",
       "   1210: Star Wars: Episode VI - Return of the Jedi (1983)   \n",
       "0                                                NaN        \\\n",
       "1                                                NaN         \n",
       "2                                                NaN         \n",
       "3                                                NaN         \n",
       "4                                                NaN         \n",
       "\n",
       "   356: Forrest Gump (1994)  318: Shawshank Redemption, The (1994)   \n",
       "0                       NaN                                    NaN  \\\n",
       "1                       NaN                                    NaN   \n",
       "2                       NaN                                    NaN   \n",
       "3                       NaN                                    NaN   \n",
       "4                       NaN                                    NaN   \n",
       "\n",
       "   593: Silence of the Lambs, The (1991)  3578: Gladiator (2000)   \n",
       "0                                    NaN                     NaN  \\\n",
       "1                                    NaN                     NaN   \n",
       "2                                    NaN                     NaN   \n",
       "3                                    NaN                     NaN   \n",
       "4                                    NaN                     NaN   \n",
       "\n",
       "   260: Star Wars: Episode IV - A New Hope (1977)   \n",
       "0                                             NaN  \\\n",
       "1                                             NaN   \n",
       "2                                             NaN   \n",
       "3                                             NaN   \n",
       "4                                             NaN   \n",
       "\n",
       "   2028: Saving Private Ryan (1998)  296: Pulp Fiction (1994)  ...   \n",
       "0                               NaN                       NaN  ...  \\\n",
       "1                               NaN                       NaN  ...   \n",
       "2                               NaN                       NaN  ...   \n",
       "3                               NaN                       NaN  ...   \n",
       "4                               NaN                       NaN  ...   \n",
       "\n",
       "   2396: Shakespeare in Love (1998)  2916: Total Recall (1990)   \n",
       "0                               NaN                        NaN  \\\n",
       "1                               NaN                        NaN   \n",
       "2                               NaN                        NaN   \n",
       "3                               NaN                        NaN   \n",
       "4                               NaN                        NaN   \n",
       "\n",
       "   780: Independence Day (ID4) (1996)  541: Blade Runner (1982)   \n",
       "0                                 NaN                       NaN  \\\n",
       "1                                 NaN                       NaN   \n",
       "2                                 NaN                       NaN   \n",
       "3                                 NaN                       NaN   \n",
       "4                                 NaN                       NaN   \n",
       "\n",
       "   1265: Groundhog Day (1993)  2571: Matrix, The (1999)   \n",
       "0                         NaN                       NaN  \\\n",
       "1                         NaN                       NaN   \n",
       "2                         NaN                       NaN   \n",
       "3                         NaN                       NaN   \n",
       "4                         NaN                       NaN   \n",
       "\n",
       "   527: Schindler's List (1993)  2762: Sixth Sense, The (1999)   \n",
       "0                           NaN                            NaN  \\\n",
       "1                           NaN                            NaN   \n",
       "2                           NaN                            NaN   \n",
       "3                           NaN                            NaN   \n",
       "4                           NaN                            NaN   \n",
       "\n",
       "   1198: Raiders of the Lost Ark (1981)  34: Babe (1995)  \n",
       "0                                   NaN              NaN  \n",
       "1                                   NaN              NaN  \n",
       "2                                   NaN              NaN  \n",
       "3                                   NaN              NaN  \n",
       "4                                   NaN              NaN  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matrix.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "matrix = pd.DataFrame(cosine_similarity(data.values[:-1,1:-1].T),index=matrix.index,columns=matrix.columns[1:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "matrix = matrix.applymap(lambda x:max(0,x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['1: Toy Story (1995)',\n",
       "       '1210: Star Wars: Episode VI - Return of the Jedi (1983)',\n",
       "       '356: Forrest Gump (1994)', '318: Shawshank Redemption, The (1994)',\n",
       "       '593: Silence of the Lambs, The (1991)', '3578: Gladiator (2000)',\n",
       "       '260: Star Wars: Episode IV - A New Hope (1977)',\n",
       "       '2028: Saving Private Ryan (1998)', '296: Pulp Fiction (1994)',\n",
       "       '1259: Stand by Me (1986)', '2396: Shakespeare in Love (1998)',\n",
       "       '2916: Total Recall (1990)', '780: Independence Day (ID4) (1996)',\n",
       "       '541: Blade Runner (1982)', '1265: Groundhog Day (1993)',\n",
       "       '2571: Matrix, The (1999)', '527: Schindler's List (1993)',\n",
       "       '2762: Sixth Sense, The (1999)', '1198: Raiders of the Lost Ark (1981)',\n",
       "       '34: Babe (1995)'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matrix.columns\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "260: Star Wars: Episode IV - A New Hope (1977)    0.747409\n",
       "780: Independence Day (ID4) (1996)                0.690665\n",
       "296: Pulp Fiction (1994)                          0.667846\n",
       "318: Shawshank Redemption, The (1994)             0.667424\n",
       "1265: Groundhog Day (1993)                        0.661016\n",
       "Name: 0, dtype: float64"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matrix.iloc[0].nlargest(6).iloc[1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_score(row,user):\n",
    "    user_rating = data.loc[(data.User==user)]\n",
    "    user_hist_item = user_rating.columns[pd.notnull(user_rating).values[0]]\n",
    "    movie_name = row.name\n",
    "\n",
    "    neighbor_names = user_hist_item.tolist()#row.index.tolist()\n",
    "\n",
    "    if 'User' in neighbor_names:\n",
    "        neighbor_names.remove('User')\n",
    "    if 'Mean' in neighbor_names:\n",
    "        neighbor_names.remove('Mean')\n",
    "    a = row.loc[neighbor_names].values\n",
    "    b = data.loc[data.User==user,neighbor_names]\n",
    "\n",
    "    return np.dot(a,b.values[0])/np.sum(a) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "user_rating = data.loc[data.User==5277]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['1: Toy Story (1995)',\n",
       " '1210: Star Wars: Episode VI - Return of the Jedi (1983)',\n",
       " '356: Forrest Gump (1994)',\n",
       " '318: Shawshank Redemption, The (1994)',\n",
       " '593: Silence of the Lambs, The (1991)',\n",
       " '3578: Gladiator (2000)',\n",
       " '260: Star Wars: Episode IV - A New Hope (1977)',\n",
       " '2028: Saving Private Ryan (1998)',\n",
       " '296: Pulp Fiction (1994)',\n",
       " '1259: Stand by Me (1986)',\n",
       " '2396: Shakespeare in Love (1998)',\n",
       " '2916: Total Recall (1990)',\n",
       " '780: Independence Day (ID4) (1996)',\n",
       " '541: Blade Runner (1982)',\n",
       " '1265: Groundhog Day (1993)',\n",
       " '2571: Matrix, The (1999)',\n",
       " \"527: Schindler's List (1993)\",\n",
       " '2762: Sixth Sense, The (1999)',\n",
       " '1198: Raiders of the Lost Ark (1981)',\n",
       " '34: Babe (1995)']"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "idx = user_rating.columns[pd.notnull(user_rating).values[0]].tolist()\n",
    "idx.remove('User')\n",
    "idx.remove('Mean')\n",
    "\n",
    "idx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "ans = matrix.apply(get_score,axis=1,args=(5277,))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Content-Based "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "ml_notes",
   "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": "3.11.2"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}