Matrix_Factorization.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/OmidGhadami95/ALS_vs_GradientDescent/blob/main/Matrix_Factorization.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Gradient Descent**"
      ],
      "metadata": {
        "id": "DmwQKHB2p8vh"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import csv\n",
        "\n",
        "count = 0\n",
        "\n",
        "with open('train.csv', 'r') as file:\n",
        "    csv_reader = csv.reader(file)\n",
        "    next(csv_reader)  # Skip header row if present\n",
        "\n",
        "    for row in csv_reader:\n",
        "        if row[0] == '13':  # Assuming user-id is the first column\n",
        "            count += 1\n",
        "\n",
        "print(f\"Number of rows with user-id '13': {count}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "at-u2XCgLce6",
        "outputId": "ce7cb837-a369-48ba-d007-82394f198928"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of rows with user-id '13': 641\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import csv\n",
        "\n",
        "unique_rows = set()\n",
        "\n",
        "with open('train.csv', 'r') as file:\n",
        "    csv_reader = csv.reader(file)\n",
        "    next(csv_reader)  # Skip header row if present\n",
        "\n",
        "    for row in csv_reader:\n",
        "        unique_rows.add(tuple(row))  # Convert row to tuple and add to set\n",
        "\n",
        "print(f\"Number of distinct records: {len(unique_rows)}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Rbwp74mnOFGf",
        "outputId": "c0ff358c-8d74-43cc-ae5c-abae76247f11"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of distinct records: 65717\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import csv\n",
        "\n",
        "distinct_user_ids = {}  # Using a dictionary instead of a set\n",
        "\n",
        "with open('train.csv', 'r') as file:\n",
        "    csv_reader = csv.reader(file)\n",
        "    next(csv_reader)  # Skip header row if present\n",
        "\n",
        "    for row in csv_reader:\n",
        "        user_id = row[0]  # Assuming user-id is the first column\n",
        "        distinct_user_ids[user_id] = True  # Using the user_id as a key in the dictionary\n",
        "\n",
        "print(f\"Number of distinct user-ids in train.csv (excluding index): {len(distinct_user_ids)}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0AZHGdc6Oxu4",
        "outputId": "5655b9c9-d7f0-4c8c-896c-a76a15710cb7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of distinct user-ids in train.csv (excluding index): 306\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "# Step 1: Load the data\n",
        "df = pd.read_csv('train.csv')\n",
        "\n",
        "# Step 2: Split the data (90% for training, 10% for validation)\n",
        "train_df, val_df = train_test_split(df, test_size=0.1, random_state=42)\n",
        "\n",
        "# Step 3: Save the training and validation sets to separate CSV files\n",
        "train_df.to_csv('train_split.csv', index=False)\n",
        "val_df.to_csv('validation_split.csv', index=False)\n",
        "\n",
        "print(\"Training and validation datasets created successfully!\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cVG5siTjPber",
        "outputId": "9f358d9d-ff22-4d12-b5a1-39019693aae6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training and validation datasets created successfully!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "# Step 1: Load the data\n",
        "df = pd.read_csv('train.csv')\n",
        "\n",
        "# Step 2: Split the data (90% for training, 10% for validation)\n",
        "train_df, val_df = train_test_split(df, test_size=0.1, random_state=42)\n",
        "\n",
        "# Step 3: Find out how many records each CSV has\n",
        "print(f\"Training set records: {len(train_df)}\")\n",
        "print(f\"Validation set records: {len(val_df)}\")\n",
        "\n",
        "# Step 4: Save the training and validation sets to separate CSV files\n",
        "train_df.to_csv('train_split.csv', index=False)\n",
        "val_df.to_csv('validation_split.csv', index=False)\n",
        "\n",
        "print(\"Training and validation datasets created successfully!\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BiA_0_UsQxMh",
        "outputId": "e1a3c02e-6efb-4d4c-bad1-899f9d67de8a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training set records: 59145\n",
            "Validation set records: 6572\n",
            "Training and validation datasets created successfully!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "\n",
        "# Load the validation split data\n",
        "val_df = pd.read_csv('validation_split.csv')\n",
        "\n",
        "# Drop the third column (assuming it's the column at index 2)\n",
        "val_df = val_df.drop(val_df.columns[2], axis=1)\n",
        "\n",
        "# Save the new DataFrame without the third column to a new CSV file\n",
        "val_df.to_csv('validation_split_no_third_column.csv', index=False)\n",
        "\n",
        "print(\"New validation_split_no_third_column.csv created without the third column!\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9-bp94l5Soxm",
        "outputId": "ef36d9a1-609b-4ab4-f023-1d35575a8075"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "New validation_split_no_third_column.csv created without the third column!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Step 1: Load the Data\n",
        "train_data = pd.read_csv('/content/train_split.csv')\n",
        "validation_data = pd.read_csv('/content/validation_split.csv')\n",
        "train_data.columns = ['userId', 'movieId', 'rating']\n",
        "validation_data.columns = ['userId', 'movieId', 'rating']\n",
        "\n",
        "# Get the number of users and movies\n",
        "num_users = train_data['userId'].nunique()\n",
        "num_movies = train_data['movieId'].nunique()\n",
        "\n",
        "# Create mappings for user and movie IDs\n",
        "user_id_map = {id: i for i, id in enumerate(train_data['userId'].unique())}\n",
        "movie_id_map = {id: i for i, id in enumerate(train_data['movieId'].unique())}\n",
        "\n",
        "# Create ratings matrix\n",
        "ratings_matrix = np.zeros((num_users, num_movies))\n",
        "for _, row in train_data.iterrows():\n",
        "    user = user_id_map[row['userId']]\n",
        "    movie = movie_id_map[row['movieId']]\n",
        "    ratings_matrix[user, movie] = row['rating']\n",
        "\n",
        "# Center the ratings\n",
        "user_means = np.mean(ratings_matrix, axis=1, where=ratings_matrix != 0)\n",
        "movie_means = np.mean(ratings_matrix, axis=0, where=ratings_matrix != 0)\n",
        "global_mean = np.mean(ratings_matrix[ratings_matrix != 0])\n",
        "\n",
        "# Create centered ratings matrix\n",
        "centered_ratings_matrix = np.zeros_like(ratings_matrix)\n",
        "for i in range(num_users):\n",
        "    for j in range(num_movies):\n",
        "        if ratings_matrix[i, j] != 0:\n",
        "            centered_ratings_matrix[i, j] = ratings_matrix[i, j] - user_means[i] - movie_means[j] + global_mean\n",
        "\n",
        "# Update train_data with centered ratings\n",
        "train_data['centered_rating'] = train_data.apply(lambda row:\n",
        "    centered_ratings_matrix[user_id_map[row['userId']], movie_id_map[row['movieId']]], axis=1)\n",
        "\n",
        "# Update validation_data with centered ratings\n",
        "validation_data['centered_rating'] = validation_data.apply(lambda row:\n",
        "    row['rating'] - user_means[user_id_map.get(row['userId'], 0)] -\n",
        "    movie_means[movie_id_map.get(row['movieId'], 0)] + global_mean\n",
        "    if row['userId'] in user_id_map and row['movieId'] in movie_id_map else 0, axis=1)\n",
        "\n",
        "# Step 2: Initialize matrices for latent factors\n",
        "num_factors = 100\n",
        "np.random.seed(42)\n",
        "user_factors = np.random.normal(scale=0.1, size=(num_users, num_factors))\n",
        "movie_factors = np.random.normal(scale=0.1, size=(num_movies, num_factors))\n",
        "\n",
        "# Step 3: Define the Matrix Factorization model\n",
        "def predict(user, movie):\n",
        "    if user in user_id_map and movie in movie_id_map:\n",
        "        return np.dot(user_factors[user_id_map[user]], movie_factors[movie_id_map[movie]])\n",
        "    return global_mean  # Return global mean for unknown user-movie pairs\n",
        "\n",
        "# Step 4: Define the loss function (Mean Absolute Error)\n",
        "def compute_loss(data):\n",
        "    predictions = np.array([predict(row['userId'], row['movieId']) for _, row in data.iterrows()])\n",
        "    return np.mean(np.abs(predictions - data['centered_rating']))\n",
        "\n",
        "# Step 5: Implement gradient descent with dynamic learning rate and gradient clipping\n",
        "initial_learning_rate = 0.1  # Reduced learning rate for MAE\n",
        "min_learning_rate = 0.001\n",
        "decay_factor = 0.7\n",
        "decay_steps = 7  # Number of epochs after which to decay the learning rate\n",
        "\n",
        "regularization = 0.2\n",
        "epochs = 100\n",
        "batch_size = 2000\n",
        "\n",
        "# Lists to store loss values and learning rates\n",
        "train_losses = []\n",
        "val_losses = []\n",
        "learning_rates = []\n",
        "\n",
        "def get_learning_rate(epoch):\n",
        "    return max(initial_learning_rate * (decay_factor ** (epoch // decay_steps)), min_learning_rate)\n",
        "\n",
        "def clip_gradient(gradient, max_norm=1.0):\n",
        "    norm = np.linalg.norm(gradient)\n",
        "    if norm > max_norm:\n",
        "        return gradient * (max_norm / norm)\n",
        "    return gradient\n",
        "\n",
        "for epoch in range(epochs):\n",
        "    learning_rate = get_learning_rate(epoch)\n",
        "    learning_rates.append(learning_rate)\n",
        "\n",
        "    for start in range(0, len(train_data), batch_size):\n",
        "        end = start + batch_size\n",
        "        batch = train_data.iloc[start:end]\n",
        "\n",
        "        for _, row in batch.iterrows():\n",
        "            user, movie, rating = row['userId'], row['movieId'], row['centered_rating']\n",
        "            prediction = predict(user, movie)\n",
        "            error = prediction - rating\n",
        "\n",
        "            # Compute gradients (sign of error for MAE)\n",
        "            gradient_direction = np.sign(error)\n",
        "            user_gradient = gradient_direction * movie_factors[movie_id_map[movie]] + regularization * user_factors[user_id_map[user]]\n",
        "            movie_gradient = gradient_direction * user_factors[user_id_map[user]] + regularization * movie_factors[movie_id_map[movie]]\n",
        "\n",
        "            # Clip gradients\n",
        "            user_gradient = clip_gradient(user_gradient)\n",
        "            movie_gradient = clip_gradient(movie_gradient)\n",
        "\n",
        "            # Update latent factors\n",
        "            user_factors[user_id_map[user]] -= learning_rate * user_gradient\n",
        "            movie_factors[movie_id_map[movie]] -= learning_rate * movie_gradient\n",
        "\n",
        "    # Compute and store loss for every epoch\n",
        "    train_loss = compute_loss(train_data)\n",
        "    val_loss = compute_loss(validation_data)\n",
        "    train_losses.append(train_loss)\n",
        "    val_losses.append(val_loss)\n",
        "    print(f\"Epoch {epoch+1}/{epochs}, LR: {learning_rate:.6f}, Train Loss (MAE): {train_loss:.4f}, Validation Loss (MAE): {val_loss:.4f}\")\n",
        "\n",
        "# Step 6: Make predictions on validation set\n",
        "validation_data['predicted_rating'] = validation_data.apply(lambda row:\n",
        "    predict(row['userId'], row['movieId']) +\n",
        "    user_means[user_id_map.get(row['userId'], 0)] +\n",
        "    movie_means[movie_id_map.get(row['movieId'], 0)] -\n",
        "    global_mean, axis=1)\n",
        "\n",
        "# Step 7: Compute final MAE on validation set\n",
        "mae = np.mean(np.abs(validation_data['predicted_rating'] - validation_data['rating']))\n",
        "print(f\"Final MAE on validation set: {mae:.4f}\")\n",
        "\n",
        "# Step 8: Visualize results\n",
        "plt.figure(figsize=(10, 6))\n",
        "plt.scatter(validation_data['rating'], validation_data['predicted_rating'], alpha=0.1)\n",
        "plt.plot([validation_data['rating'].min(), validation_data['rating'].max()],\n",
        "         [validation_data['rating'].min(), validation_data['rating'].max()],\n",
        "         'r--', lw=2)\n",
        "plt.xlabel('Actual Ratings')\n",
        "plt.ylabel('Predicted Ratings')\n",
        "plt.title('Actual vs Predicted Ratings')\n",
        "plt.show()\n",
        "\n",
        "# Step 9: Plot training and validation loss\n",
        "plt.figure(figsize=(10, 6))\n",
        "epochs_range = range(1, epochs + 1)\n",
        "plt.plot(epochs_range, train_losses, label='Training Loss (MAE)')\n",
        "plt.plot(epochs_range, val_losses, label='Validation Loss (MAE)')\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Loss')\n",
        "plt.title('Training and Validation Loss based on MAE')\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "# Step 10: Plot learning rate decay\n",
        "plt.figure(figsize=(10, 6))\n",
        "plt.plot(range(1, epochs + 1), learning_rates)\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Learning Rate')\n",
        "plt.title('Learning Rate Decay')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "Ug-D2ChxkpBO",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "dbcd15bc-fec9-41df-8eb3-6e22eec76d5d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100, LR: 0.100000, Train Loss (MAE): 0.6304, Validation Loss (MAE): 0.6724\n",
            "Epoch 2/100, LR: 0.100000, Train Loss (MAE): 0.6299, Validation Loss (MAE): 0.6722\n",
            "Epoch 3/100, LR: 0.100000, Train Loss (MAE): 0.6283, Validation Loss (MAE): 0.6717\n",
            "Epoch 4/100, LR: 0.100000, Train Loss (MAE): 0.6258, Validation Loss (MAE): 0.6710\n",
            "Epoch 5/100, LR: 0.100000, Train Loss (MAE): 0.6223, Validation Loss (MAE): 0.6700\n",
            "Epoch 6/100, LR: 0.100000, Train Loss (MAE): 0.6179, Validation Loss (MAE): 0.6687\n",
            "Epoch 7/100, LR: 0.100000, Train Loss (MAE): 0.6131, Validation Loss (MAE): 0.6675\n",
            "Epoch 8/100, LR: 0.070000, Train Loss (MAE): 0.6071, Validation Loss (MAE): 0.6659\n",
            "Epoch 9/100, LR: 0.070000, Train Loss (MAE): 0.6027, Validation Loss (MAE): 0.6651\n",
            "Epoch 10/100, LR: 0.070000, Train Loss (MAE): 0.5983, Validation Loss (MAE): 0.6643\n",
            "Epoch 11/100, LR: 0.070000, Train Loss (MAE): 0.5938, Validation Loss (MAE): 0.6635\n",
            "Epoch 12/100, LR: 0.070000, Train Loss (MAE): 0.5892, Validation Loss (MAE): 0.6626\n",
            "Epoch 13/100, LR: 0.070000, Train Loss (MAE): 0.5853, Validation Loss (MAE): 0.6623\n",
            "Epoch 14/100, LR: 0.070000, Train Loss (MAE): 0.5809, Validation Loss (MAE): 0.6616\n",
            "Epoch 15/100, LR: 0.049000, Train Loss (MAE): 0.5749, Validation Loss (MAE): 0.6606\n",
            "Epoch 16/100, LR: 0.049000, Train Loss (MAE): 0.5710, Validation Loss (MAE): 0.6601\n",
            "Epoch 17/100, LR: 0.049000, Train Loss (MAE): 0.5673, Validation Loss (MAE): 0.6596\n",
            "Epoch 18/100, LR: 0.049000, Train Loss (MAE): 0.5637, Validation Loss (MAE): 0.6591\n",
            "Epoch 19/100, LR: 0.049000, Train Loss (MAE): 0.5604, Validation Loss (MAE): 0.6591\n",
            "Epoch 20/100, LR: 0.049000, Train Loss (MAE): 0.5568, Validation Loss (MAE): 0.6586\n",
            "Epoch 21/100, LR: 0.049000, Train Loss (MAE): 0.5535, Validation Loss (MAE): 0.6583\n",
            "Epoch 22/100, LR: 0.034300, Train Loss (MAE): 0.5484, Validation Loss (MAE): 0.6576\n",
            "Epoch 23/100, LR: 0.034300, Train Loss (MAE): 0.5451, Validation Loss (MAE): 0.6573\n",
            "Epoch 24/100, LR: 0.034300, Train Loss (MAE): 0.5422, Validation Loss (MAE): 0.6572\n",
            "Epoch 25/100, LR: 0.034300, Train Loss (MAE): 0.5396, Validation Loss (MAE): 0.6570\n",
            "Epoch 26/100, LR: 0.034300, Train Loss (MAE): 0.5368, Validation Loss (MAE): 0.6566\n",
            "Epoch 27/100, LR: 0.034300, Train Loss (MAE): 0.5340, Validation Loss (MAE): 0.6567\n",
            "Epoch 28/100, LR: 0.034300, Train Loss (MAE): 0.5319, Validation Loss (MAE): 0.6566\n",
            "Epoch 29/100, LR: 0.024010, Train Loss (MAE): 0.5277, Validation Loss (MAE): 0.6560\n",
            "Epoch 30/100, LR: 0.024010, Train Loss (MAE): 0.5250, Validation Loss (MAE): 0.6560\n",
            "Epoch 31/100, LR: 0.024010, Train Loss (MAE): 0.5231, Validation Loss (MAE): 0.6558\n",
            "Epoch 32/100, LR: 0.024010, Train Loss (MAE): 0.5207, Validation Loss (MAE): 0.6557\n",
            "Epoch 33/100, LR: 0.024010, Train Loss (MAE): 0.5187, Validation Loss (MAE): 0.6556\n",
            "Epoch 34/100, LR: 0.024010, Train Loss (MAE): 0.5166, Validation Loss (MAE): 0.6554\n",
            "Epoch 35/100, LR: 0.024010, Train Loss (MAE): 0.5145, Validation Loss (MAE): 0.6553\n",
            "Epoch 36/100, LR: 0.016807, Train Loss (MAE): 0.5121, Validation Loss (MAE): 0.6551\n",
            "Epoch 37/100, LR: 0.016807, Train Loss (MAE): 0.5100, Validation Loss (MAE): 0.6550\n",
            "Epoch 38/100, LR: 0.016807, Train Loss (MAE): 0.5081, Validation Loss (MAE): 0.6550\n",
            "Epoch 39/100, LR: 0.016807, Train Loss (MAE): 0.5067, Validation Loss (MAE): 0.6549\n",
            "Epoch 40/100, LR: 0.016807, Train Loss (MAE): 0.5050, Validation Loss (MAE): 0.6549\n",
            "Epoch 41/100, LR: 0.016807, Train Loss (MAE): 0.5037, Validation Loss (MAE): 0.6548\n",
            "Epoch 42/100, LR: 0.016807, Train Loss (MAE): 0.5022, Validation Loss (MAE): 0.6547\n",
            "Epoch 43/100, LR: 0.011765, Train Loss (MAE): 0.5000, Validation Loss (MAE): 0.6545\n",
            "Epoch 44/100, LR: 0.011765, Train Loss (MAE): 0.4986, Validation Loss (MAE): 0.6546\n",
            "Epoch 45/100, LR: 0.011765, Train Loss (MAE): 0.4974, Validation Loss (MAE): 0.6545\n",
            "Epoch 46/100, LR: 0.011765, Train Loss (MAE): 0.4960, Validation Loss (MAE): 0.6546\n",
            "Epoch 47/100, LR: 0.011765, Train Loss (MAE): 0.4951, Validation Loss (MAE): 0.6544\n",
            "Epoch 48/100, LR: 0.011765, Train Loss (MAE): 0.4940, Validation Loss (MAE): 0.6545\n",
            "Epoch 49/100, LR: 0.011765, Train Loss (MAE): 0.4930, Validation Loss (MAE): 0.6545\n",
            "Epoch 50/100, LR: 0.008235, Train Loss (MAE): 0.4914, Validation Loss (MAE): 0.6543\n",
            "Epoch 51/100, LR: 0.008235, Train Loss (MAE): 0.4904, Validation Loss (MAE): 0.6543\n",
            "Epoch 52/100, LR: 0.008235, Train Loss (MAE): 0.4896, Validation Loss (MAE): 0.6543\n",
            "Epoch 53/100, LR: 0.008235, Train Loss (MAE): 0.4887, Validation Loss (MAE): 0.6543\n",
            "Epoch 54/100, LR: 0.008235, Train Loss (MAE): 0.4878, Validation Loss (MAE): 0.6542\n",
            "Epoch 55/100, LR: 0.008235, Train Loss (MAE): 0.4871, Validation Loss (MAE): 0.6542\n",
            "Epoch 56/100, LR: 0.008235, Train Loss (MAE): 0.4863, Validation Loss (MAE): 0.6542\n",
            "Epoch 57/100, LR: 0.005765, Train Loss (MAE): 0.4855, Validation Loss (MAE): 0.6542\n",
            "Epoch 58/100, LR: 0.005765, Train Loss (MAE): 0.4846, Validation Loss (MAE): 0.6542\n",
            "Epoch 59/100, LR: 0.005765, Train Loss (MAE): 0.4839, Validation Loss (MAE): 0.6541\n",
            "Epoch 60/100, LR: 0.005765, Train Loss (MAE): 0.4834, Validation Loss (MAE): 0.6542\n",
            "Epoch 61/100, LR: 0.005765, Train Loss (MAE): 0.4829, Validation Loss (MAE): 0.6541\n",
            "Epoch 62/100, LR: 0.005765, Train Loss (MAE): 0.4822, Validation Loss (MAE): 0.6541\n",
            "Epoch 63/100, LR: 0.005765, Train Loss (MAE): 0.4817, Validation Loss (MAE): 0.6541\n",
            "Epoch 64/100, LR: 0.004035, Train Loss (MAE): 0.4811, Validation Loss (MAE): 0.6541\n",
            "Epoch 65/100, LR: 0.004035, Train Loss (MAE): 0.4805, Validation Loss (MAE): 0.6541\n",
            "Epoch 66/100, LR: 0.004035, Train Loss (MAE): 0.4801, Validation Loss (MAE): 0.6541\n",
            "Epoch 67/100, LR: 0.004035, Train Loss (MAE): 0.4797, Validation Loss (MAE): 0.6541\n",
            "Epoch 68/100, LR: 0.004035, Train Loss (MAE): 0.4792, Validation Loss (MAE): 0.6541\n",
            "Epoch 69/100, LR: 0.004035, Train Loss (MAE): 0.4788, Validation Loss (MAE): 0.6541\n",
            "Epoch 70/100, LR: 0.004035, Train Loss (MAE): 0.4784, Validation Loss (MAE): 0.6541\n",
            "Epoch 71/100, LR: 0.002825, Train Loss (MAE): 0.4780, Validation Loss (MAE): 0.6540\n",
            "Epoch 72/100, LR: 0.002825, Train Loss (MAE): 0.4776, Validation Loss (MAE): 0.6540\n",
            "Epoch 73/100, LR: 0.002825, Train Loss (MAE): 0.4773, Validation Loss (MAE): 0.6540\n",
            "Epoch 74/100, LR: 0.002825, Train Loss (MAE): 0.4770, Validation Loss (MAE): 0.6540\n",
            "Epoch 75/100, LR: 0.002825, Train Loss (MAE): 0.4767, Validation Loss (MAE): 0.6540\n",
            "Epoch 76/100, LR: 0.002825, Train Loss (MAE): 0.4764, Validation Loss (MAE): 0.6540\n",
            "Epoch 77/100, LR: 0.002825, Train Loss (MAE): 0.4762, Validation Loss (MAE): 0.6540\n",
            "Epoch 78/100, LR: 0.001977, Train Loss (MAE): 0.4758, Validation Loss (MAE): 0.6540\n",
            "Epoch 79/100, LR: 0.001977, Train Loss (MAE): 0.4755, Validation Loss (MAE): 0.6540\n",
            "Epoch 80/100, LR: 0.001977, Train Loss (MAE): 0.4753, Validation Loss (MAE): 0.6540\n",
            "Epoch 81/100, LR: 0.001977, Train Loss (MAE): 0.4751, Validation Loss (MAE): 0.6540\n",
            "Epoch 82/100, LR: 0.001977, Train Loss (MAE): 0.4749, Validation Loss (MAE): 0.6540\n",
            "Epoch 83/100, LR: 0.001977, Train Loss (MAE): 0.4747, Validation Loss (MAE): 0.6540\n",
            "Epoch 84/100, LR: 0.001977, Train Loss (MAE): 0.4745, Validation Loss (MAE): 0.6540\n",
            "Epoch 85/100, LR: 0.001384, Train Loss (MAE): 0.4743, Validation Loss (MAE): 0.6540\n",
            "Epoch 86/100, LR: 0.001384, Train Loss (MAE): 0.4741, Validation Loss (MAE): 0.6540\n",
            "Epoch 87/100, LR: 0.001384, Train Loss (MAE): 0.4739, Validation Loss (MAE): 0.6540\n",
            "Epoch 88/100, LR: 0.001384, Train Loss (MAE): 0.4738, Validation Loss (MAE): 0.6540\n",
            "Epoch 89/100, LR: 0.001384, Train Loss (MAE): 0.4736, Validation Loss (MAE): 0.6540\n",
            "Epoch 90/100, LR: 0.001384, Train Loss (MAE): 0.4735, Validation Loss (MAE): 0.6539\n",
            "Epoch 91/100, LR: 0.001384, Train Loss (MAE): 0.4733, Validation Loss (MAE): 0.6540\n",
            "Epoch 92/100, LR: 0.001000, Train Loss (MAE): 0.4732, Validation Loss (MAE): 0.6539\n",
            "Epoch 93/100, LR: 0.001000, Train Loss (MAE): 0.4731, Validation Loss (MAE): 0.6539\n",
            "Epoch 94/100, LR: 0.001000, Train Loss (MAE): 0.4729, Validation Loss (MAE): 0.6539\n",
            "Epoch 95/100, LR: 0.001000, Train Loss (MAE): 0.4728, Validation Loss (MAE): 0.6539\n",
            "Epoch 96/100, LR: 0.001000, Train Loss (MAE): 0.4727, Validation Loss (MAE): 0.6540\n",
            "Epoch 97/100, LR: 0.001000, Train Loss (MAE): 0.4726, Validation Loss (MAE): 0.6539\n",
            "Epoch 98/100, LR: 0.001000, Train Loss (MAE): 0.4725, Validation Loss (MAE): 0.6539\n",
            "Epoch 99/100, LR: 0.001000, Train Loss (MAE): 0.4724, Validation Loss (MAE): 0.6539\n",
            "Epoch 100/100, LR: 0.001000, Train Loss (MAE): 0.4723, Validation Loss (MAE): 0.6539\n",
            "Final MAE on validation set: 0.6539\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Step 8: Save predictions to CSV\n",
        "prediction_df = validation_data[['userId', 'movieId', 'rating', 'predicted_rating']]\n",
        "prediction_df.to_csv('predicted_ratings.csv', index=False)\n",
        "print(\"Predictions saved to 'predicted_ratings.csv'\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bfXG59TfoqXW",
        "outputId": "37b1ab7c-238d-44d2-ded7-e60ac04a5fd2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Predictions saved to 'predicted_ratings.csv'\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "from sklearn.metrics import mean_absolute_error, r2_score\n",
        "\n",
        "# Step 1: Load the CSV files\n",
        "prediction_df = pd.read_csv('predicted_ratings.csv')\n",
        "validation_df = pd.read_csv('/content/validation_split.csv')\n",
        "\n",
        "# Step 2: Extract the predicted and actual values (third column in each file)\n",
        "predicted_values = prediction_df.iloc[:, 3]  # Third column\n",
        "actual_values = validation_df.iloc[:, 2]     # Third column\n",
        "\n",
        "# Step 3: Calculate loss (using mean absolute error for regression tasks)\n",
        "mse_loss = mean_absolute_error(actual_values, predicted_values)\n",
        "print(f\"Mean Absolute Error (Loss): {mse_loss:.4f}\")\n",
        "\n",
        "# Step 4: Calculate R-squared score (another metric for regression tasks)\n",
        "r2 = r2_score(actual_values, predicted_values)\n",
        "print(f\"R-squared: {r2:.4f}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "264603d1-6abb-4766-ae6f-eaabdc2f94ea",
        "id": "67NHFZYVowqx"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mean Absolute Error (Loss): 0.6539\n",
            "R-squared: 0.3060\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Step 7: Load the new test data\n",
        "new_test_file = '/content/test_without_rating.csv'\n",
        "new_test_data = pd.read_csv(new_test_file)\n",
        "new_test_data.columns = ['userId', 'movieId']\n",
        "\n",
        "# Step 8: Predict ratings for the new test set\n",
        "new_test_data['predicted_rating'] = new_test_data.apply(lambda row:\n",
        "    predict(row['userId'], row['movieId']) +\n",
        "    user_means[user_id_map.get(row['userId'], 0)] +\n",
        "    movie_means[movie_id_map.get(row['movieId'], 0)] -\n",
        "    global_mean, axis=1)\n",
        "\n",
        "# Step 9: Create a new DataFrame with predictions and save to CSV\n",
        "new_test_predictions_df = new_test_data.copy()\n",
        "\n",
        "# Round the predicted ratings to 3 decimal places\n",
        "new_test_predictions_df['predicted_rating'] = new_test_predictions_df['predicted_rating'].round(3)\n",
        "\n",
        "# Ensure predicted ratings are within the valid range (0.5 to 5)\n",
        "new_test_predictions_df['predicted_rating'] = new_test_predictions_df['predicted_rating'].clip(0.5, 5)\n",
        "\n",
        "# Save the predictions to a new CSV file\n",
        "new_test_predictions_df.to_csv('test_predictions_gradient_mae.csv', index=False)\n",
        "print(\"New test predictions saved to 'test_predictions_gradient_mae.csv'.\")\n",
        "\n",
        "# Step 10: Display the first few rows of the predictions\n",
        "print(new_test_predictions_df.head())\n",
        "\n",
        "# Step 11: Provide some statistics about the predictions\n",
        "print(\"\\nPrediction Statistics:\")\n",
        "print(new_test_predictions_df['predicted_rating'].describe())\n",
        "\n",
        "# Step 12: Save the final predictions to a TXT file\n",
        "with open('predictions_gradient.txt', 'w') as f:\n",
        "    for _, row in new_test_predictions_df.iterrows():\n",
        "        f.write(f\"{row['userId']},{row['movieId']},{row['predicted_rating']:.3f}\\n\")\n",
        "\n",
        "print(\"New test predictions saved to 'predictions_gradient.txt'.\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d9ba6ea0-577e-4da6-89f9-1c57126123d2",
        "id": "7JeiNrvtkJmL"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "New test predictions saved to 'test_predictions_gradient_mae.csv'.\n",
            "   userId  movieId  predicted_rating\n",
            "0     202     1204             3.853\n",
            "1     103      367             3.858\n",
            "2      74      698             4.219\n",
            "3       2     1935             2.970\n",
            "4      49     2220             3.562\n",
            "\n",
            "Prediction Statistics:\n",
            "count    7302.000000\n",
            "mean        3.585570\n",
            "std         0.695461\n",
            "min         0.500000\n",
            "25%         3.149000\n",
            "50%         3.652000\n",
            "75%         4.080000\n",
            "max         5.000000\n",
            "Name: predicted_rating, dtype: float64\n",
            "New test predictions saved to 'predictions_gradient.txt'.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Step 7: Load the new test data\n",
        "new_test_file = '/content/test_without_rating.csv'\n",
        "new_test_data = pd.read_csv(new_test_file)\n",
        "new_test_data.columns = ['userId', 'movieId']\n",
        "\n",
        "# Step 8: Predict ratings for the new test set\n",
        "new_test_data['predicted_rating'] = new_test_data.apply(lambda row:\n",
        "    predict(row['userId'], row['movieId']) +\n",
        "    user_means[user_id_map.get(row['userId'], 0)] +\n",
        "    movie_means[movie_id_map.get(row['movieId'], 0)] -\n",
        "    global_mean, axis=1)\n",
        "\n",
        "# Step 9: Create a new DataFrame with predictions and save to CSV\n",
        "new_test_predictions_df = new_test_data.copy()\n",
        "\n",
        "# Round the predicted ratings to 3 decimal places\n",
        "new_test_predictions_df['predicted_rating'] = new_test_predictions_df['predicted_rating'].round(3)\n",
        "\n",
        "# Ensure predicted ratings are within the valid range (0.5 to 5)\n",
        "new_test_predictions_df['predicted_rating'] = new_test_predictions_df['predicted_rating'].clip(0.5, 5)\n",
        "\n",
        "# Save the full predictions to a CSV file\n",
        "new_test_predictions_df.to_csv('test_predictions_gradient_mae.csv', index=False)\n",
        "print(\"Full test predictions saved to 'test_predictions_gradient_mae.csv'.\")\n",
        "\n",
        "# Save only the predicted ratings to a new CSV file\n",
        "predicted_ratings_only = new_test_predictions_df['predicted_rating']\n",
        "predicted_ratings_only.to_csv('predicted_ratings_only.csv', index=False, header=False)\n",
        "print(\"Predicted ratings only saved to 'predicted_ratings_only.csv'.\")\n",
        "\n",
        "# Step 10: Display the first few rows of the predictions\n",
        "print(new_test_predictions_df.head())\n",
        "\n",
        "# Step 11: Provide some statistics about the predictions\n",
        "print(\"\\nPrediction Statistics:\")\n",
        "print(new_test_predictions_df['predicted_rating'].describe())\n",
        "\n",
        "# Step 12: Save the final full predictions to a TXT file\n",
        "with open('predictions_gradient.txt', 'w') as f:\n",
        "    for _, row in new_test_predictions_df.iterrows():\n",
        "        f.write(f\"{row['userId']},{row['movieId']},{row['predicted_rating']:.3f}\\n\")\n",
        "\n",
        "print(\"Full test predictions saved to 'predictions_gradient.txt'.\")\n",
        "\n",
        "# Save only the predicted ratings to a new TXT file\n",
        "with open('Gradient_predicted_ratings_only.txt', 'w') as f:\n",
        "    for rating in new_test_predictions_df['predicted_rating']:\n",
        "        f.write(f\"{rating:.3f}\\n\")\n",
        "\n",
        "print(\"Predicted ratings only saved to 'Gradient_predicted_ratings_only.txt'.\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xr2wzhCiBUOt",
        "outputId": "67d41710-ca2b-4630-ed31-c0f5e6b65a67"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Full test predictions saved to 'test_predictions_gradient_mae.csv'.\n",
            "Predicted ratings only saved to 'predicted_ratings_only.csv'.\n",
            "   userId  movieId  predicted_rating\n",
            "0     202     1204             3.853\n",
            "1     103      367             3.858\n",
            "2      74      698             4.219\n",
            "3       2     1935             2.970\n",
            "4      49     2220             3.562\n",
            "\n",
            "Prediction Statistics:\n",
            "count    7302.000000\n",
            "mean        3.585570\n",
            "std         0.695461\n",
            "min         0.500000\n",
            "25%         3.149000\n",
            "50%         3.652000\n",
            "75%         4.080000\n",
            "max         5.000000\n",
            "Name: predicted_rating, dtype: float64\n",
            "Full test predictions saved to 'predictions_gradient.txt'.\n",
            "Predicted ratings only saved to 'Gradient_predicted_ratings_only.txt'.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **ALS**"
      ],
      "metadata": {
        "id": "E_9A5-YqqYV7"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "def prepare_datasets(training_file, validation_file):\n",
        "    \"\"\"\n",
        "    Prepare training and validation datasets from CSV files.\n",
        "    \"\"\"\n",
        "    training_set = pd.read_csv(training_file)\n",
        "    validation_set = pd.read_csv(validation_file)\n",
        "    training_set.columns = ['user', 'item', 'score']\n",
        "    validation_set.columns = ['user', 'item', 'score']\n",
        "\n",
        "    # Get the number of users and items\n",
        "    user_count = max(training_set['user'].max(), validation_set['user'].max()) + 1\n",
        "    item_count = max(training_set['item'].max(), validation_set['item'].max()) + 1\n",
        "\n",
        "    # Create mappings for user and item IDs\n",
        "    user_index = {id: i for i, id in enumerate(training_set['user'].unique())}\n",
        "    item_index = {id: i for i, id in enumerate(training_set['item'].unique())}\n",
        "\n",
        "    # Create the score matrix S\n",
        "    S = np.zeros((user_count, item_count))\n",
        "    for row in training_set.itertuples():\n",
        "        S[user_index[row.user], item_index[row.item]] = row.score\n",
        "\n",
        "    # Center the scores\n",
        "    user_averages = np.mean(S, axis=1, where=S != 0)\n",
        "    item_averages = np.mean(S, axis=0, where=S != 0)\n",
        "    overall_average = np.mean(S[S != 0])\n",
        "\n",
        "    # Create centered score matrix\n",
        "    centered_S = np.zeros_like(S)\n",
        "    for i in range(user_count):\n",
        "        for j in range(item_count):\n",
        "            if S[i, j] != 0:\n",
        "                centered_S[i, j] = S[i, j] - user_averages[i] - item_averages[j] + overall_average\n",
        "\n",
        "    # Update training_set with centered scores\n",
        "    training_set['centered_score'] = training_set.apply(lambda row:\n",
        "        centered_S[user_index[row['user']], item_index[row['item']]], axis=1)\n",
        "\n",
        "    # Update validation_set with centered scores\n",
        "    validation_set['centered_score'] = validation_set.apply(lambda row:\n",
        "        row['score'] - user_averages[user_index.get(row['user'], 0)] -\n",
        "        item_averages[item_index.get(row['item'], 0)] + overall_average\n",
        "        if row['user'] in user_index and row['item'] in item_index else 0, axis=1)\n",
        "\n",
        "    return centered_S, training_set, validation_set, user_count, item_count, user_averages, item_averages, overall_average, user_index, item_index\n",
        "\n",
        "def estimate_score(P, Q, data, user_averages, item_averages, overall_average, user_index, item_index):\n",
        "    \"\"\"\n",
        "    Estimate scores for the user-item pairs in data.\n",
        "    \"\"\"\n",
        "    estimations = []\n",
        "    for row in data.itertuples():\n",
        "        user, item = row.user, row.item\n",
        "        if user in user_index and item in item_index:\n",
        "            user_idx, item_idx = user_index[user], item_index[item]\n",
        "            score_est = np.dot(P[user_idx, :], Q[item_idx, :])\n",
        "            # Add back the user and item averages, and subtract the overall average\n",
        "            score_est += user_averages[user_idx] + item_averages[item_idx] - overall_average\n",
        "        else:\n",
        "            score_est = overall_average\n",
        "        estimations.append(score_est)\n",
        "\n",
        "    return np.array(estimations)\n",
        "\n",
        "def matrix_factorization(S, validation_set, latent_factors=200, lambda_reg=0.0001, iterations=50, user_index=None, item_index=None):\n",
        "    \"\"\"\n",
        "    Matrix factorization using Alternating Least Squares (ALS) with MAE loss.\n",
        "    \"\"\"\n",
        "    user_count, item_count = S.shape\n",
        "\n",
        "    # Initialize user and item matrices with random values\n",
        "    P = np.random.normal(scale=1./latent_factors, size=(user_count, latent_factors))\n",
        "    Q = np.random.normal(scale=1./latent_factors, size=(item_count, latent_factors))\n",
        "\n",
        "    # Create binary matrices to indicate observed scores\n",
        "    observed_S = S.copy()\n",
        "    observed_S[observed_S != 0] = 1\n",
        "\n",
        "    # Store the MAE for each iteration\n",
        "    error_history = []\n",
        "\n",
        "    for iteration in range(iterations):\n",
        "        # Update user latent factors\n",
        "        for i in range(user_count):\n",
        "            Q_i = Q[observed_S[i, :] > 0]\n",
        "            S_i = S[i, observed_S[i, :] > 0]\n",
        "            A = np.dot(Q_i.T, Q_i) + lambda_reg * np.eye(latent_factors)\n",
        "            b = np.dot(Q_i.T, S_i)\n",
        "            P[i] = np.linalg.solve(A, b)\n",
        "\n",
        "        # Update item latent factors\n",
        "        for j in range(item_count):\n",
        "            P_j = P[observed_S[:, j] > 0]\n",
        "            S_j = S[observed_S[:, j] > 0, j]\n",
        "            A = np.dot(P_j.T, P_j) + lambda_reg * np.eye(latent_factors)\n",
        "            b = np.dot(P_j.T, S_j)\n",
        "            Q[j] = np.linalg.solve(A, b)\n",
        "\n",
        "        # Calculate MAE on validation set\n",
        "        validation_estimations = estimate_score(P, Q, validation_set, user_averages, item_averages, overall_average, user_index, item_index)\n",
        "        mae = np.mean(np.abs(validation_set['score'] - validation_estimations))\n",
        "        error_history.append(mae)\n",
        "\n",
        "        print(f'Iteration {iteration + 1}/{iterations}, MAE: {mae:.4f}')\n",
        "\n",
        "    return P, Q, error_history\n",
        "\n",
        "def visualize_error(error_history, filename='mae_convergence_mf.png'):\n",
        "    \"\"\"\n",
        "    Visualize the convergence of the MAE.\n",
        "    \"\"\"\n",
        "    plt.figure(figsize=(10, 6))\n",
        "    plt.plot(range(len(error_history)), error_history, marker='o', linestyle='-')\n",
        "    plt.xlabel('Epoch')\n",
        "    plt.ylabel('Loss')\n",
        "    plt.title('Convergence of Loss ALS (Matrix Factorization)')\n",
        "    plt.grid(True)\n",
        "    plt.savefig(filename)\n",
        "    plt.show()\n",
        "\n",
        "# Step 1: Prepare the data\n",
        "training_file = '/content/train_split.csv'  # Replace with your training file path\n",
        "validation_file = '/content/validation_split.csv'    # Replace with your validation file path\n",
        "S, training_set, validation_set, user_count, item_count, user_averages, item_averages, overall_average, user_index, item_index = prepare_datasets(training_file, validation_file)\n",
        "\n",
        "# Step 2: Apply matrix factorization to find latent factors\n",
        "P, Q, error_history = matrix_factorization(S, validation_set, latent_factors=200, lambda_reg=0.0001, iterations=50, user_index=user_index, item_index=item_index)\n",
        "\n",
        "# Step 3: Visualize the convergence of the MAE\n",
        "visualize_error(error_history)\n",
        "\n",
        "# Step 4: Estimate scores for the validation set\n",
        "final_estimations = estimate_score(P, Q, validation_set, user_averages, item_averages, overall_average, user_index, item_index)\n",
        "\n",
        "# Step 5: Calculate final MAE on validation set\n",
        "final_mae = np.mean(np.abs(validation_set['score'] - final_estimations))\n",
        "print(f\"Final MAE on validation set: {final_mae:.4f}\")\n",
        "\n",
        "# Step 6: Save estimations to a file\n",
        "estimations_df = pd.DataFrame({\n",
        "    'user': validation_set['user'],\n",
        "    'item': validation_set['item'],\n",
        "    'estimated_score': final_estimations\n",
        "})\n",
        "estimations_df.to_csv('estimations_als.csv', index=False)\n",
        "print(\"Estimations saved to 'estimations_als.csv'.\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "71geSPeBfiEt",
        "outputId": "e21f635c-a115-4ee1-ce2b-f50b67827751"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Iteration 1/50, MAE: 0.7795\n",
            "Iteration 2/50, MAE: 0.7540\n",
            "Iteration 3/50, MAE: 0.7424\n",
            "Iteration 4/50, MAE: 0.7345\n",
            "Iteration 5/50, MAE: 0.7288\n",
            "Iteration 6/50, MAE: 0.7244\n",
            "Iteration 7/50, MAE: 0.7208\n",
            "Iteration 8/50, MAE: 0.7180\n",
            "Iteration 9/50, MAE: 0.7158\n",
            "Iteration 10/50, MAE: 0.7139\n",
            "Iteration 11/50, MAE: 0.7123\n",
            "Iteration 12/50, MAE: 0.7108\n",
            "Iteration 13/50, MAE: 0.7095\n",
            "Iteration 14/50, MAE: 0.7083\n",
            "Iteration 15/50, MAE: 0.7073\n",
            "Iteration 16/50, MAE: 0.7064\n",
            "Iteration 17/50, MAE: 0.7056\n",
            "Iteration 18/50, MAE: 0.7049\n",
            "Iteration 19/50, MAE: 0.7042\n",
            "Iteration 20/50, MAE: 0.7035\n",
            "Iteration 21/50, MAE: 0.7029\n",
            "Iteration 22/50, MAE: 0.7024\n",
            "Iteration 23/50, MAE: 0.7019\n",
            "Iteration 24/50, MAE: 0.7015\n",
            "Iteration 25/50, MAE: 0.7011\n",
            "Iteration 26/50, MAE: 0.7007\n",
            "Iteration 27/50, MAE: 0.7003\n",
            "Iteration 28/50, MAE: 0.7000\n",
            "Iteration 29/50, MAE: 0.6997\n",
            "Iteration 30/50, MAE: 0.6995\n",
            "Iteration 31/50, MAE: 0.6992\n",
            "Iteration 32/50, MAE: 0.6990\n",
            "Iteration 33/50, MAE: 0.6988\n",
            "Iteration 34/50, MAE: 0.6986\n",
            "Iteration 35/50, MAE: 0.6984\n",
            "Iteration 36/50, MAE: 0.6982\n",
            "Iteration 37/50, MAE: 0.6980\n",
            "Iteration 38/50, MAE: 0.6979\n",
            "Iteration 39/50, MAE: 0.6977\n",
            "Iteration 40/50, MAE: 0.6976\n",
            "Iteration 41/50, MAE: 0.6975\n",
            "Iteration 42/50, MAE: 0.6974\n",
            "Iteration 43/50, MAE: 0.6973\n",
            "Iteration 44/50, MAE: 0.6972\n",
            "Iteration 45/50, MAE: 0.6971\n",
            "Iteration 46/50, MAE: 0.6971\n",
            "Iteration 47/50, MAE: 0.6970\n",
            "Iteration 48/50, MAE: 0.6970\n",
            "Iteration 49/50, MAE: 0.6969\n",
            "Iteration 50/50, MAE: 0.6969\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Final MAE on validation set: 0.6969\n",
            "Estimations saved to 'estimations_als.csv'.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Prediction Function**"
      ],
      "metadata": {
        "id": "2IszgNkhcQXT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Step 7: Load the new test data\n",
        "new_test_file = '/content/test_without_rating.csv'\n",
        "new_test_data = pd.read_csv(new_test_file)\n",
        "new_test_data.columns = ['user', 'item']\n",
        "\n",
        "# Step 8: Predict ratings for the new test set\n",
        "new_test_predictions = estimate_score(P, Q, new_test_data, user_averages, item_averages, overall_average, user_index, item_index)\n",
        "\n",
        "# Step 9: Create a new DataFrame with predictions and save to CSV\n",
        "new_test_predictions_df = pd.DataFrame({\n",
        "    'user': new_test_data['user'],\n",
        "    'item': new_test_data['item'],\n",
        "    'predicted_rating': new_test_predictions\n",
        "})\n",
        "\n",
        "# Round the predicted ratings to 3 decimal places\n",
        "new_test_predictions_df['predicted_rating'] = new_test_predictions_df['predicted_rating'].round(3)\n",
        "\n",
        "# Ensure predicted ratings are within the valid range (0.5 to 5)\n",
        "new_test_predictions_df['predicted_rating'] = new_test_predictions_df['predicted_rating'].clip(0.5, 5)\n",
        "\n",
        "# Step 10: Save the predictions to a CSV file\n",
        "new_test_predictions_df.to_csv('test_predictions_mf_mae.csv', index=False)\n",
        "print(\"New test predictions saved to 'test_predictions_mf_mae.csv'.\")\n",
        "\n",
        "# Step 11: Print summary statistics of the predictions\n",
        "print(new_test_predictions_df['predicted_rating'].describe())\n",
        "\n",
        "# Step 12: Save the final predictions to a TXT file\n",
        "with open('predictions_als.txt', 'w') as f:\n",
        "    for _, row in new_test_predictions_df.iterrows():\n",
        "        f.write(f\"{row['user']},{row['item']},{row['predicted_rating']:.3f}\\n\")\n",
        "\n",
        "print(\"New test predictions saved to 'predictions_mf_mae.txt'.\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wc8qTRctbMfY",
        "outputId": "fef43200-0af8-42b3-cee2-7053ec75434d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "New test predictions saved to 'test_predictions_mf_mae.csv'.\n",
            "count    7302.000000\n",
            "mean        3.579222\n",
            "std         0.760208\n",
            "min         0.500000\n",
            "25%         3.093000\n",
            "50%         3.631000\n",
            "75%         4.118750\n",
            "max         5.000000\n",
            "Name: predicted_rating, dtype: float64\n",
            "New test predictions saved to 'predictions_mf_mae.txt'.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Step 7: Load the new test data\n",
        "new_test_file = '/content/test_without_rating.csv'\n",
        "new_test_data = pd.read_csv(new_test_file)\n",
        "new_test_data.columns = ['user', 'item']\n",
        "\n",
        "# Step 8: Predict ratings for the new test set\n",
        "new_test_predictions = estimate_score(P, Q, new_test_data, user_averages, item_averages, overall_average, user_index, item_index)\n",
        "\n",
        "# Step 9: Create a new DataFrame with predictions and save to CSV\n",
        "new_test_predictions_df = pd.DataFrame({\n",
        "    'user': new_test_data['user'],\n",
        "    'item': new_test_data['item'],\n",
        "    'predicted_rating': new_test_predictions\n",
        "})\n",
        "\n",
        "# Round the predicted ratings to 3 decimal places\n",
        "new_test_predictions_df['predicted_rating'] = new_test_predictions_df['predicted_rating'].round(3)\n",
        "\n",
        "# Ensure predicted ratings are within the valid range (0.5 to 5)\n",
        "new_test_predictions_df['predicted_rating'] = new_test_predictions_df['predicted_rating'].clip(0.5, 5)\n",
        "\n",
        "# Step 10: Save the full predictions to a CSV file\n",
        "new_test_predictions_df.to_csv('test_predictions_mf_mae.csv', index=False)\n",
        "print(\"Full test predictions saved to 'test_predictions_mf_mae.csv'.\")\n",
        "\n",
        "# Save only the predicted ratings to a new CSV file\n",
        "predicted_ratings_only = new_test_predictions_df['predicted_rating']\n",
        "predicted_ratings_only.to_csv('predicted_ratings_only.csv', index=False, header=False)\n",
        "print(\"Predicted ratings only saved to 'predicted_ratings_only.csv'.\")\n",
        "\n",
        "# Step 11: Print summary statistics of the predictions\n",
        "print(new_test_predictions_df['predicted_rating'].describe())\n",
        "\n",
        "# Step 12: Save the final full predictions to a TXT file\n",
        "with open('predictions_mf_mae.txt', 'w') as f:\n",
        "    for _, row in new_test_predictions_df.iterrows():\n",
        "        f.write(f\"{row['user']},{row['item']},{row['predicted_rating']:.3f}\\n\")\n",
        "\n",
        "print(\"Full test predictions saved to 'predictions_mf_mae.txt'.\")\n",
        "\n",
        "# Save only the predicted ratings to a new TXT file\n",
        "with open('ALS_predicted_ratings_only.txt', 'w') as f:\n",
        "    for rating in new_test_predictions_df['predicted_rating']:\n",
        "        f.write(f\"{rating:.3f}\\n\")\n",
        "\n",
        "print(\"Predicted ratings only saved to 'ALS_predicted_ratings_only.txt'.\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VRApdROXBnU_",
        "outputId": "cac25541-7b33-4e64-f531-a34c114985cd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Full test predictions saved to 'test_predictions_mf_mae.csv'.\n",
            "Predicted ratings only saved to 'predicted_ratings_only.csv'.\n",
            "count    7302.000000\n",
            "mean        3.579222\n",
            "std         0.760208\n",
            "min         0.500000\n",
            "25%         3.093000\n",
            "50%         3.631000\n",
            "75%         4.118750\n",
            "max         5.000000\n",
            "Name: predicted_rating, dtype: float64\n",
            "Full test predictions saved to 'predictions_mf_mae.txt'.\n",
            "Predicted ratings only saved to 'ALS_predicted_ratings_only.txt'.\n"
          ]
        }
      ]
    }
  ]
}