{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "50cda6d2-d65a-4b39-b35c-cd00d1d5c10b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===== Dataset 1 OVERVIEW =====\n",
      "\n",
      "Shape:\n",
      "Rows: 2096\n",
      "Columns: 17\n",
      "\n",
      "Column types:\n",
      "ID                               object\n",
      "Date                             object\n",
      "SCORE                           float64\n",
      "ForwardHead                     float64\n",
      "LeftArmFallForward              float64\n",
      "RightArmFallForward             float64\n",
      "LeftShoulderElevation           float64\n",
      "RightShoulderElevation          float64\n",
      "ExcessiveForwardLean            float64\n",
      "LeftAsymmetricalWeightShift     float64\n",
      "RightAsymmetricalWeightShift    float64\n",
      "LeftKneeMovesInward             float64\n",
      "RightKneeMovesInward            float64\n",
      "LeftKneeMovesOutward            float64\n",
      "RightKneeMovesOutward           float64\n",
      "LeftHeelRises                   float64\n",
      "RightHeelRises                  float64\n",
      "dtype: object\n",
      "\n",
      "Missing values per column:\n",
      "ID                              0\n",
      "Date                            0\n",
      "SCORE                           0\n",
      "ForwardHead                     0\n",
      "LeftArmFallForward              0\n",
      "RightArmFallForward             0\n",
      "LeftShoulderElevation           0\n",
      "RightShoulderElevation          0\n",
      "ExcessiveForwardLean            0\n",
      "LeftAsymmetricalWeightShift     0\n",
      "RightAsymmetricalWeightShift    0\n",
      "LeftKneeMovesInward             0\n",
      "RightKneeMovesInward            0\n",
      "LeftKneeMovesOutward            0\n",
      "RightKneeMovesOutward           0\n",
      "LeftHeelRises                   0\n",
      "RightHeelRises                  0\n",
      "dtype: int64\n",
      "\n",
      "Basic statistics:\n",
      "                                                 ID       Date        SCORE  \\\n",
      "count                                          2096       2096  2096.000000   \n",
      "unique                                         2096         92          NaN   \n",
      "top     cf0e95ab-d6ec-475a-9cc0-c9444b8dd080.Kinect  9/11/2018          NaN   \n",
      "freq                                              1        131          NaN   \n",
      "mean                                            NaN        NaN     0.572420   \n",
      "std                                             NaN        NaN     0.236856   \n",
      "min                                             NaN        NaN     0.010000   \n",
      "25%                                             NaN        NaN     0.407843   \n",
      "50%                                             NaN        NaN     0.617248   \n",
      "75%                                             NaN        NaN     0.759416   \n",
      "max                                             NaN        NaN     0.993987   \n",
      "\n",
      "        ForwardHead  LeftArmFallForward  RightArmFallForward  \\\n",
      "count   2096.000000         2096.000000          2096.000000   \n",
      "unique          NaN                 NaN                  NaN   \n",
      "top             NaN                 NaN                  NaN   \n",
      "freq            NaN                 NaN                  NaN   \n",
      "mean       0.301579            0.996842             0.880792   \n",
      "std        0.477732            0.481902             0.578039   \n",
      "min        0.000000            0.000000             0.000000   \n",
      "25%        0.000000            0.690000             0.170000   \n",
      "50%        0.000000            1.120000             1.110000   \n",
      "75%        0.480000            1.390000             1.420000   \n",
      "max        1.500000            1.500000             1.500000   \n",
      "\n",
      "        LeftShoulderElevation  RightShoulderElevation  ExcessiveForwardLean  \\\n",
      "count             2096.000000             2096.000000           2096.000000   \n",
      "unique                    NaN                     NaN                   NaN   \n",
      "top                       NaN                     NaN                   NaN   \n",
      "freq                      NaN                     NaN                   NaN   \n",
      "mean                 0.271293                0.465000              0.201150   \n",
      "std                  0.418118                0.475603              0.419338   \n",
      "min                  0.000000                0.000000              0.000000   \n",
      "25%                  0.000000                0.030000              0.000000   \n",
      "50%                  0.030000                0.325000              0.000000   \n",
      "75%                  0.390000                0.852500              0.080000   \n",
      "max                  1.500000                1.500000              1.500000   \n",
      "\n",
      "        LeftAsymmetricalWeightShift  RightAsymmetricalWeightShift  \\\n",
      "count                   2096.000000                   2096.000000   \n",
      "unique                          NaN                           NaN   \n",
      "top                             NaN                           NaN   \n",
      "freq                            NaN                           NaN   \n",
      "mean                       0.175219                      0.065315   \n",
      "std                        0.397579                      0.252260   \n",
      "min                        0.000000                      0.000000   \n",
      "25%                        0.000000                      0.000000   \n",
      "50%                        0.000000                      0.000000   \n",
      "75%                        0.050000                      0.000000   \n",
      "max                        1.500000                      1.500000   \n",
      "\n",
      "        LeftKneeMovesInward  RightKneeMovesInward  LeftKneeMovesOutward  \\\n",
      "count           2096.000000           2096.000000           2096.000000   \n",
      "unique                  NaN                   NaN                   NaN   \n",
      "top                     NaN                   NaN                   NaN   \n",
      "freq                    NaN                   NaN                   NaN   \n",
      "mean               0.037657              0.179089              0.170787   \n",
      "std                0.140765              0.329131              0.401551   \n",
      "min                0.000000              0.000000              0.000000   \n",
      "25%                0.000000              0.000000              0.000000   \n",
      "50%                0.000000              0.000000              0.000000   \n",
      "75%                0.000000              0.200000              0.020000   \n",
      "max                1.310000              1.500000              1.500000   \n",
      "\n",
      "        RightKneeMovesOutward  LeftHeelRises  RightHeelRises  \n",
      "count             2096.000000    2096.000000     2096.000000  \n",
      "unique                    NaN            NaN             NaN  \n",
      "top                       NaN            NaN             NaN  \n",
      "freq                      NaN            NaN             NaN  \n",
      "mean                 0.238912       0.059241        0.102505  \n",
      "std                  0.528886       0.226736        0.290971  \n",
      "min                  0.000000       0.000000        0.000000  \n",
      "25%                  0.000000       0.000000        0.000000  \n",
      "50%                  0.000000       0.000000        0.000000  \n",
      "75%                  0.000000       0.000000        0.000000  \n",
      "max                  1.500000       1.500000        1.500000  \n",
      "\n",
      "Low-variance features (possible under-sampled):\n",
      "Series([], dtype: int64)\n",
      "\n",
      "Categorical value distribution:\n",
      "\n",
      "Column: ID\n",
      "ID\n",
      "cf0e95ab-d6ec-475a-9cc0-c9444b8dd080.Kinect    0.000477\n",
      "b0cf6cce-156f-42f1-8f04-775d7556233f.Kinect    0.000477\n",
      "4072bb99-8aa7-4cab-b4de-986865ab47f1.Kinect    0.000477\n",
      "8452bac0-735d-4eae-8dab-01827e37481d.Kinect    0.000477\n",
      "5d668b1e-3be1-48cf-836b-f9d327c7b4bf.Kinect    0.000477\n",
      "                                                 ...   \n",
      "28a8db8e-d663-4485-a3d1-e35913e8d08b.Kinect    0.000477\n",
      "315a5309-5f24-4c26-a291-739a1ec0bd25.Kinect    0.000477\n",
      "5d0810f4-6ecc-41bd-83ba-f512176ddf55.Kinect    0.000477\n",
      "c86cbd71-962b-4885-96e3-a960bd1e89b9.Kinect    0.000477\n",
      "8f1b9646-337a-4692-a501-78010c314b4a.Kinect    0.000477\n",
      "Name: proportion, Length: 2096, dtype: float64\n",
      "\n",
      "Column: Date\n",
      "Date\n",
      "9/11/2018     0.062500\n",
      "9/26/2018     0.054389\n",
      "7/20/2018     0.042462\n",
      "8/23/2018     0.041031\n",
      "5/10/2018     0.030534\n",
      "                ...   \n",
      "6/18/2018     0.000477\n",
      "8/3/2018      0.000477\n",
      "8/10/2018     0.000477\n",
      "10/14/2018    0.000477\n",
      "8/12/2018     0.000477\n",
      "Name: proportion, Length: 92, dtype: float64\n",
      "\n",
      "===== Dataset 2 OVERVIEW =====\n",
      "\n",
      "Shape:\n",
      "Rows: 2094\n",
      "Columns: 43\n",
      "\n",
      "Column types:\n",
      "AimoScore                float64\n",
      "No_1_Angle_Deviation     float64\n",
      "No_2_Angle_Deviation     float64\n",
      "No_3_Angle_Deviation     float64\n",
      "No_4_Angle_Deviation     float64\n",
      "No_5_Angle_Deviation     float64\n",
      "No_6_Angle_Deviation     float64\n",
      "No_7_Angle_Deviation     float64\n",
      "No_8_Angle_Deviation     float64\n",
      "No_9_Angle_Deviation     float64\n",
      "No_10_Angle_Deviation    float64\n",
      "No_11_Angle_Deviation    float64\n",
      "No_12_Angle_Deviation    float64\n",
      "No_13_Angle_Deviation    float64\n",
      "No_1_NASM_Deviation      float64\n",
      "No_2_NASM_Deviation      float64\n",
      "No_3_NASM_Deviation      float64\n",
      "No_4_NASM_Deviation      float64\n",
      "No_5_NASM_Deviation      float64\n",
      "No_6_NASM_Deviation      float64\n",
      "No_7_NASM_Deviation      float64\n",
      "No_8_NASM_Deviation      float64\n",
      "No_9_NASM_Deviation      float64\n",
      "No_10_NASM_Deviation     float64\n",
      "No_11_NASM_Deviation     float64\n",
      "No_12_NASM_Deviation     float64\n",
      "No_13_NASM_Deviation     float64\n",
      "No_14_NASM_Deviation     float64\n",
      "No_15_NASM_Deviation     float64\n",
      "No_16_NASM_Deviation     float64\n",
      "No_17_NASM_Deviation     float64\n",
      "No_18_NASM_Deviation     float64\n",
      "No_19_NASM_Deviation     float64\n",
      "No_20_NASM_Deviation     float64\n",
      "No_21_NASM_Deviation     float64\n",
      "No_22_NASM_Deviation     float64\n",
      "No_23_NASM_Deviation     float64\n",
      "No_24_NASM_Deviation     float64\n",
      "No_25_NASM_Deviation     float64\n",
      "No_1_Time_Deviation      float64\n",
      "No_2_Time_Deviation      float64\n",
      "EstimatedScore           float64\n",
      "ID                        object\n",
      "dtype: object\n",
      "\n",
      "Missing values per column:\n",
      "AimoScore                0\n",
      "No_1_Angle_Deviation     0\n",
      "No_2_Angle_Deviation     0\n",
      "No_3_Angle_Deviation     0\n",
      "No_4_Angle_Deviation     0\n",
      "No_5_Angle_Deviation     0\n",
      "No_6_Angle_Deviation     0\n",
      "No_7_Angle_Deviation     0\n",
      "No_8_Angle_Deviation     0\n",
      "No_9_Angle_Deviation     0\n",
      "No_10_Angle_Deviation    0\n",
      "No_11_Angle_Deviation    0\n",
      "No_12_Angle_Deviation    0\n",
      "No_13_Angle_Deviation    0\n",
      "No_1_NASM_Deviation      0\n",
      "No_2_NASM_Deviation      0\n",
      "No_3_NASM_Deviation      0\n",
      "No_4_NASM_Deviation      0\n",
      "No_5_NASM_Deviation      0\n",
      "No_6_NASM_Deviation      0\n",
      "No_7_NASM_Deviation      0\n",
      "No_8_NASM_Deviation      0\n",
      "No_9_NASM_Deviation      0\n",
      "No_10_NASM_Deviation     0\n",
      "No_11_NASM_Deviation     0\n",
      "No_12_NASM_Deviation     0\n",
      "No_13_NASM_Deviation     0\n",
      "No_14_NASM_Deviation     0\n",
      "No_15_NASM_Deviation     0\n",
      "No_16_NASM_Deviation     0\n",
      "No_17_NASM_Deviation     0\n",
      "No_18_NASM_Deviation     0\n",
      "No_19_NASM_Deviation     0\n",
      "No_20_NASM_Deviation     0\n",
      "No_21_NASM_Deviation     0\n",
      "No_22_NASM_Deviation     0\n",
      "No_23_NASM_Deviation     0\n",
      "No_24_NASM_Deviation     0\n",
      "No_25_NASM_Deviation     0\n",
      "No_1_Time_Deviation      0\n",
      "No_2_Time_Deviation      0\n",
      "EstimatedScore           0\n",
      "ID                       0\n",
      "dtype: int64\n",
      "\n",
      "Basic statistics:\n",
      "          AimoScore  No_1_Angle_Deviation  No_2_Angle_Deviation  \\\n",
      "count   2094.000000           2094.000000           2094.000000   \n",
      "unique          NaN                   NaN                   NaN   \n",
      "top             NaN                   NaN                   NaN   \n",
      "freq            NaN                   NaN                   NaN   \n",
      "mean       0.572740              0.539248              0.504462   \n",
      "std        0.236510              0.239284              0.282136   \n",
      "min        0.010000              0.279770              0.092300   \n",
      "25%        0.407907              0.279770              0.250149   \n",
      "50%        0.617248              0.500000              0.500000   \n",
      "75%        0.759378              0.749761              0.750239   \n",
      "max        0.993987              1.000000              1.000000   \n",
      "\n",
      "        No_3_Angle_Deviation  No_4_Angle_Deviation  No_5_Angle_Deviation  \\\n",
      "count            2094.000000           2094.000000           2094.000000   \n",
      "unique                   NaN                   NaN                   NaN   \n",
      "top                      NaN                   NaN                   NaN   \n",
      "freq                     NaN                   NaN                   NaN   \n",
      "mean                0.560258              0.518065              0.513932   \n",
      "std                 0.217828              0.264104              0.270076   \n",
      "min                 0.346724              0.188905              0.163080   \n",
      "25%                 0.346724              0.250239              0.250717   \n",
      "50%                 0.500000              0.500000              0.500478   \n",
      "75%                 0.750239              0.750239              0.750239   \n",
      "max                 1.000000              1.000000              1.000000   \n",
      "\n",
      "        No_6_Angle_Deviation  No_7_Angle_Deviation  No_8_Angle_Deviation  \\\n",
      "count            2094.000000           2094.000000           2094.000000   \n",
      "unique                   NaN                   NaN                   NaN   \n",
      "top                      NaN                   NaN                   NaN   \n",
      "freq                     NaN                   NaN                   NaN   \n",
      "mean                0.510336              0.519434              0.635923   \n",
      "std                 0.274473              0.262884              0.153280   \n",
      "min                 0.140124              0.194165              0.303247   \n",
      "25%                 0.250717              0.250717              0.521760   \n",
      "50%                 0.500478              0.500956              0.521760   \n",
      "75%                 0.750239              0.750239              0.749761   \n",
      "max                 1.000000              1.000000              1.000000   \n",
      "\n",
      "        No_9_Angle_Deviation  ...  No_20_NASM_Deviation  No_21_NASM_Deviation  \\\n",
      "count            2094.000000  ...           2094.000000           2094.000000   \n",
      "unique                   NaN  ...                   NaN                   NaN   \n",
      "top                      NaN  ...                   NaN                   NaN   \n",
      "freq                     NaN  ...                   NaN                   NaN   \n",
      "mean                0.512548  ...              0.715943              0.706319   \n",
      "std                 0.271193  ...              0.100437              0.105886   \n",
      "min                 0.156385  ...              0.656624              0.640401   \n",
      "25%                 0.250717  ...              0.656624              0.642276   \n",
      "50%                 0.500478  ...              0.656624              0.642276   \n",
      "75%                 0.750239  ...              0.751196              0.749761   \n",
      "max                 1.000000  ...              1.000000              1.000000   \n",
      "\n",
      "        No_22_NASM_Deviation  No_23_NASM_Deviation  No_24_NASM_Deviation  \\\n",
      "count            2094.000000           2094.000000           2094.000000   \n",
      "unique                   NaN                   NaN                   NaN   \n",
      "top                      NaN                   NaN                   NaN   \n",
      "freq                     NaN                   NaN                   NaN   \n",
      "mean                0.652969              0.710774              0.667517   \n",
      "std                 0.141000              0.103196              0.130989   \n",
      "min                 0.525310              0.648972              0.578192   \n",
      "25%                 0.552846              0.648972              0.578192   \n",
      "50%                 0.552846              0.648972              0.578192   \n",
      "75%                 0.749880              0.751196              0.751196   \n",
      "max                 1.000000              1.000000              1.000000   \n",
      "\n",
      "        No_25_NASM_Deviation  No_1_Time_Deviation  No_2_Time_Deviation  \\\n",
      "count            2094.000000          2094.000000          2094.000000   \n",
      "unique                   NaN                  NaN                  NaN   \n",
      "top                      NaN                  NaN                  NaN   \n",
      "freq                     NaN                  NaN                  NaN   \n",
      "mean                0.548089             0.512021             0.512373   \n",
      "std                 0.230642             0.273056             0.272647   \n",
      "min                 0.308943             0.148733             0.151124   \n",
      "25%                 0.308943             0.250717             0.250717   \n",
      "50%                 0.500000             0.500478             0.500478   \n",
      "75%                 0.750239             0.750239             0.750239   \n",
      "max                 1.000000             1.000000             1.000000   \n",
      "\n",
      "        EstimatedScore                                           ID  \n",
      "count      2094.000000                                         2094  \n",
      "unique             NaN                                         2091  \n",
      "top                NaN  ff63b8a9-deb5-4179-b424-fd07d04ca37f.Kinect  \n",
      "freq               NaN                                            2  \n",
      "mean          0.500209                                          NaN  \n",
      "std           0.288790                                          NaN  \n",
      "min           0.000478                                          NaN  \n",
      "25%           0.250239                                          NaN  \n",
      "50%           0.500478                                          NaN  \n",
      "75%           0.750239                                          NaN  \n",
      "max           1.000000                                          NaN  \n",
      "\n",
      "[11 rows x 43 columns]\n",
      "\n",
      "Low-variance features (possible under-sampled):\n",
      "Series([], dtype: int64)\n",
      "\n",
      "Categorical value distribution:\n",
      "\n",
      "Column: ID\n",
      "ID\n",
      "ff63b8a9-deb5-4179-b424-fd07d04ca37f.Kinect    0.000955\n",
      "ff514cba-6c5e-41a9-9936-3e8400565c8d.Kinect    0.000955\n",
      "ff015ddd-c9fb-49e3-8e33-abbf72e8b120.Kinect    0.000955\n",
      "a90b85c8-d284-4441-a586-1399c1525102.Kinect    0.000478\n",
      "a7cbdc66-91e7-4f63-ac9e-5ccb1c6a7580.Kinect    0.000478\n",
      "                                                 ...   \n",
      "5443e1f4-af4d-4b98-9597-eda7cbb8a71b.Kinect    0.000478\n",
      "53fa6c2e-f0c5-4e28-8b72-1ad82a23b36b.Kinect    0.000478\n",
      "53ef7e0b-d741-4556-9237-83ad75e4d1b1.Kinect    0.000478\n",
      "53e0e447-2c7c-4a5d-8f87-dc1307102990.Kinect    0.000478\n",
      "80a9b249-205d-4496-a6d9-815a5ec2029e.Kinect    0.000478\n",
      "Name: proportion, Length: 2091, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "REPO_ROOT = os.path.abspath(os.path.join(os.getcwd(), \"..\"))\n",
    "DATA_DIR = os.path.join(REPO_ROOT, \"Datasets_all\")\n",
    "# Load datasets\n",
    "df1 = pd.read_csv(os.path.join(DATA_DIR,\"scores_and_weaklink.csv\"))\n",
    "df2 = pd.read_csv(os.path.join(DATA_DIR, \"aimoscores.csv\"))   \n",
    "def dataset_overview(df, name=\"Dataset\"):\n",
    "    print(f\"\\n===== {name} OVERVIEW =====\")\n",
    "    \n",
    "    # Basic shape\n",
    "    print(\"\\nShape:\")\n",
    "    print(f\"Rows: {df.shape[0]}\")\n",
    "    print(f\"Columns: {df.shape[1]}\")\n",
    "    \n",
    "    # Column types\n",
    "    print(\"\\nColumn types:\")\n",
    "    print(df.dtypes)\n",
    "    \n",
    "    # Missing values\n",
    "    print(\"\\nMissing values per column:\")\n",
    "    print(df.isnull().sum())\n",
    "    \n",
    "    # Basic statistics\n",
    "    print(\"\\nBasic statistics:\")\n",
    "    print(df.describe(include='all'))\n",
    "    \n",
    "    # Low variance features (almost constant)\n",
    "    print(\"\\nLow-variance features (possible under-sampled):\")\n",
    "    low_variance = df.nunique()\n",
    "    print(low_variance[low_variance <= 3])\n",
    "    \n",
    "    # Categorical imbalance\n",
    "    print(\"\\nCategorical value distribution:\")\n",
    "    categorical_cols = df.select_dtypes(include=['object', 'category']).columns\n",
    "    \n",
    "    for col in categorical_cols:\n",
    "        print(f\"\\nColumn: {col}\")\n",
    "        print(df[col].value_counts(normalize=True))\n",
    "\n",
    "dataset_overview(df1, \"Dataset 1\")\n",
    "dataset_overview(df2, \"Dataset 2\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b7ae7d5c-a7a1-4fb1-a099-fe87d5066c1f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===== UNDER-SAMPLED FEATURES =====\n",
      "\n",
      "Column: ID\n",
      "Rare categories (<5%):\n",
      "ID\n",
      "cf0e95ab-d6ec-475a-9cc0-c9444b8dd080.Kinect    0.000477\n",
      "b0cf6cce-156f-42f1-8f04-775d7556233f.Kinect    0.000477\n",
      "4072bb99-8aa7-4cab-b4de-986865ab47f1.Kinect    0.000477\n",
      "8452bac0-735d-4eae-8dab-01827e37481d.Kinect    0.000477\n",
      "5d668b1e-3be1-48cf-836b-f9d327c7b4bf.Kinect    0.000477\n",
      "                                                 ...   \n",
      "28a8db8e-d663-4485-a3d1-e35913e8d08b.Kinect    0.000477\n",
      "315a5309-5f24-4c26-a291-739a1ec0bd25.Kinect    0.000477\n",
      "5d0810f4-6ecc-41bd-83ba-f512176ddf55.Kinect    0.000477\n",
      "c86cbd71-962b-4885-96e3-a960bd1e89b9.Kinect    0.000477\n",
      "8f1b9646-337a-4692-a501-78010c314b4a.Kinect    0.000477\n",
      "Name: proportion, Length: 2096, dtype: float64\n",
      "\n",
      "Column: Date\n",
      "Rare categories (<5%):\n",
      "Date\n",
      "7/20/2018     0.042462\n",
      "8/23/2018     0.041031\n",
      "5/10/2018     0.030534\n",
      "7/18/2018     0.028626\n",
      "6/19/2018     0.028149\n",
      "                ...   \n",
      "6/18/2018     0.000477\n",
      "8/3/2018      0.000477\n",
      "8/10/2018     0.000477\n",
      "10/14/2018    0.000477\n",
      "8/12/2018     0.000477\n",
      "Name: proportion, Length: 90, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "def find_under_sampled_features(df, threshold=0.05):\n",
    "    print(\"\\n===== UNDER-SAMPLED FEATURES =====\")\n",
    "    \n",
    "    categorical_cols = df.select_dtypes(include=['object', 'category']).columns\n",
    "    \n",
    "    for col in categorical_cols:\n",
    "        value_dist = df[col].value_counts(normalize=True)\n",
    "        rare_values = value_dist[value_dist < threshold]\n",
    "        \n",
    "        if not rare_values.empty:\n",
    "            print(f\"\\nColumn: {col}\")\n",
    "            print(\"Rare categories (<5%):\")\n",
    "            print(rare_values)\n",
    "find_under_sampled_features(df1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "44b6c2a4-c450-41a1-a2a3-2cb495b3e54a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x1000 with 16 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "df1.hist(figsize=(12,10))\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "563b22f0-cad7-43e9-b312-9200d1a33745",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[DS1] Train: 1676 rows | Test: 420 rows\n",
      "  [DS1 Bootstrap 1] 1676 rows saved.\n",
      "  [DS1 Bootstrap 2] 1676 rows saved.\n",
      "  [DS1 Bootstrap 3] 1676 rows saved.\n",
      "\n",
      "[DS2] Train: 1675 rows | Test: 419 rows\n",
      "  [DS2 Upsampled train] 1676 rows saved.\n",
      "  [DS2 Bootstrap 1] 1676 rows saved.\n",
      "  [DS2 Bootstrap 2] 1676 rows saved.\n",
      "  [DS2 Bootstrap 3] 1676 rows saved.\n",
      "\n",
      "── Files written to /Users/reemothman/Downloads/Data-intensive-systems/A5/data ──\n",
      "  df1_test.csv                                     420 rows\n",
      "  df1_train.csv                                   1676 rows\n",
      "  df1_train_bootstrap_1.csv                       1676 rows\n",
      "  df1_train_bootstrap_2.csv                       1676 rows\n",
      "  df1_train_bootstrap_3.csv                       1676 rows\n",
      "  df2_test.csv                                     419 rows\n",
      "  df2_train.csv                                   1675 rows\n",
      "  df2_train_bootstrap_1.csv                       1676 rows\n",
      "  df2_train_bootstrap_2.csv                       1676 rows\n",
      "  df2_train_bootstrap_3.csv                       1676 rows\n",
      "  df2_train_upsampled.csv                         1676 rows\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.utils import resample\n",
    "\n",
    "\n",
    "REPO_ROOT = os.path.abspath(os.path.join(os.getcwd(), \"..\"))\n",
    "DATA_DIR  = os.path.join(REPO_ROOT, \"Datasets_all\")\n",
    "OUT_DIR   = os.path.join(os.getcwd(), \"data\")          # existing 'data' folder next to notebook\n",
    "os.makedirs(OUT_DIR, exist_ok=True)\n",
    "\n",
    "# load\n",
    "df1 = pd.read_csv(os.path.join(DATA_DIR, \"scores_and_weaklink.csv\"))\n",
    "df2 = pd.read_csv(os.path.join(DATA_DIR, \"aimoscores.csv\"))\n",
    "\n",
    "RANDOM_STATE = 42\n",
    "\n",
    "# DATASET 1  – will be used for CLASSIFICATION (after merging with df2 on ID)\n",
    "# No class label yet, so we split now and save; upsampling will be done after\n",
    "# the merge & label creation step. We still create bootstrap resamples here.\n",
    "\n",
    "df1_train, df1_test = train_test_split(df1, test_size=0.2, random_state=RANDOM_STATE)\n",
    "\n",
    "print(f\"[DS1] Train: {len(df1_train)} rows | Test: {len(df1_test)} rows\")\n",
    "\n",
    "# Save base splits\n",
    "df1_train.to_csv(os.path.join(OUT_DIR, \"df1_train.csv\"), index=False)\n",
    "df1_test.to_csv(os.path.join(OUT_DIR,  \"df1_test.csv\"),  index=False)\n",
    "\n",
    "# Bootstrap resamples from the training set (for ensemble diversity)\n",
    "N_BOOTSTRAP = 3\n",
    "for i in range(1, N_BOOTSTRAP + 1):\n",
    "    boot = resample(df1_train, replace=True, n_samples=len(df1_train), random_state=RANDOM_STATE + i)\n",
    "    boot.to_csv(os.path.join(OUT_DIR, f\"df1_train_bootstrap_{i}.csv\"), index=False)\n",
    "    print(f\"  [DS1 Bootstrap {i}] {len(boot)} rows saved.\")\n",
    "    \n",
    "# DATASET 2  – used for REGRESSION  (target = AimoScore)\n",
    "# Upsampling for regression = bootstrap resampling (no class imbalance concept)\n",
    "\n",
    "df2_train, df2_test = train_test_split(df2, test_size=0.2, random_state=RANDOM_STATE)\n",
    "\n",
    "print(f\"\\n[DS2] Train: {len(df2_train)} rows | Test: {len(df2_test)} rows\")\n",
    "\n",
    "# Save base splits\n",
    "df2_train.to_csv(os.path.join(OUT_DIR, \"df2_train.csv\"), index=False)\n",
    "df2_test.to_csv(os.path.join(OUT_DIR,  \"df2_test.csv\"),  index=False)\n",
    "\n",
    "# ── Upsampling strategy for regression: bootstrap + score-bin-aware resampling\n",
    "# Bin AimoScore into quartiles; oversample under-represented score ranges\n",
    "df2_train = df2_train.copy()\n",
    "df2_train[\"_score_bin\"] = pd.qcut(df2_train[\"AimoScore\"], q=4, labels=False)\n",
    "\n",
    "bin_counts  = df2_train[\"_score_bin\"].value_counts()\n",
    "target_size = bin_counts.max()          # upsample all bins to match the largest\n",
    "\n",
    "upsampled_parts = []\n",
    "for bin_label, group in df2_train.groupby(\"_score_bin\"):\n",
    "    if len(group) < target_size:\n",
    "        group_up = resample(group, replace=True, n_samples=target_size, random_state=RANDOM_STATE)\n",
    "    else:\n",
    "        group_up = group\n",
    "    upsampled_parts.append(group_up)\n",
    "\n",
    "df2_train_upsampled = pd.concat(upsampled_parts).drop(columns=[\"_score_bin\"]).sample(frac=1, random_state=RANDOM_STATE)\n",
    "df2_train.drop(columns=[\"_score_bin\"], inplace=True)   # clean original too\n",
    "\n",
    "df2_train_upsampled.to_csv(os.path.join(OUT_DIR, \"df2_train_upsampled.csv\"), index=False)\n",
    "print(f\"  [DS2 Upsampled train] {len(df2_train_upsampled)} rows saved.\")\n",
    "\n",
    "# Bootstrap resamples from the upsampled training set\n",
    "for i in range(1, N_BOOTSTRAP + 1):\n",
    "    boot = resample(df2_train_upsampled, replace=True,\n",
    "                    n_samples=len(df2_train_upsampled), random_state=RANDOM_STATE + i)\n",
    "    boot.to_csv(os.path.join(OUT_DIR, f\"df2_train_bootstrap_{i}.csv\"), index=False)\n",
    "    print(f\"  [DS2 Bootstrap {i}] {len(boot)} rows saved.\")\n",
    "\n",
    "# Summary\n",
    "print(\"\\n── Files written to\", OUT_DIR, \"──\")\n",
    "for f in sorted(os.listdir(OUT_DIR)):\n",
    "    path = os.path.join(OUT_DIR, f)\n",
    "    rows = pd.read_csv(path).shape[0]\n",
    "    print(f\"  {f:45s}  {rows:>5} rows\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "1cf8b011-da73-42b8-9cb2-994cf1726c58",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasets loaded:\n",
      "Dataset 1: (2096, 17)\n",
      "Dataset 2: (2094, 43)\n",
      "Train/test split completed.\n",
      "Generating upsampled datasets for Dataset 1...\n",
      "Generating upsampled datasets for Dataset 2...\n",
      "All augmented datasets saved in /data folder.\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "\n",
    "\n",
    "REPO_ROOT = os.path.abspath(os.path.join(os.getcwd(), \"..\"))\n",
    "DATA_DIR = os.path.join(REPO_ROOT, \"Datasets_all\")\n",
    "\n",
    "OUTPUT_DIR = os.path.join(os.getcwd(), \"data\")\n",
    "os.makedirs(OUTPUT_DIR, exist_ok=True)\n",
    "\n",
    "# Load datasets\n",
    "df1 = pd.read_csv(os.path.join(DATA_DIR, \"scores_and_weaklink.csv\"))\n",
    "df2 = pd.read_csv(os.path.join(DATA_DIR, \"aimoscores.csv\"))\n",
    "\n",
    "print(\"Datasets loaded:\")\n",
    "print(\"Dataset 1:\", df1.shape)\n",
    "print(\"Dataset 2:\", df2.shape)\n",
    "\n",
    "\n",
    "\n",
    "# STEP 1 — TRAIN / TEST SPLIT \n",
    "\n",
    "\n",
    "train_df1, test_df1 = train_test_split(df1, test_size=0.2, random_state=42)\n",
    "train_df2, test_df2 = train_test_split(df2, test_size=0.2, random_state=42)\n",
    "\n",
    "# Save test sets (never touched again)\n",
    "test_df1.to_csv(os.path.join(OUTPUT_DIR, \"dataset1_test.csv\"), index=False)\n",
    "test_df2.to_csv(os.path.join(OUTPUT_DIR, \"dataset2_test.csv\"), index=False)\n",
    "\n",
    "print(\"Train/test split completed.\")\n",
    "\n",
    "# STEP 2 — UPSAMPLING FUNCTION\n",
    "# (Bootstrap + noise, ID preserved)\n",
    "\n",
    "def generate_upsampled_datasets(df, n_datasets=10, noise_level=0.01):\n",
    "\n",
    "    augmented_datasets = []\n",
    "    \n",
    "    numeric_cols = df.select_dtypes(include=np.number).columns\n",
    "    id_cols = [col for col in df.columns if col == \"ID\"]\n",
    "    non_numeric_cols = [col for col in df.columns if col not in numeric_cols and col != \"ID\"]\n",
    "    \n",
    "    for i in range(n_datasets):\n",
    "        \n",
    "        # Bootstrap sampling\n",
    "        sampled_df = df.sample(frac=1, replace=True, random_state=42 + i).reset_index(drop=True)\n",
    "        \n",
    "        # Add noise ONLY to numeric columns\n",
    "        noise = np.random.normal(0, noise_level, sampled_df[numeric_cols].shape)\n",
    "        sampled_df[numeric_cols] = sampled_df[numeric_cols] + noise\n",
    "        \n",
    "        augmented_datasets.append(sampled_df)\n",
    "    \n",
    "    return augmented_datasets\n",
    "\n",
    "\n",
    "# STEP 3 — GENERATE UPSAMPLED TRAINING DATASETS\n",
    "\n",
    "print(\"Generating upsampled datasets for Dataset 1...\")\n",
    "upsampled_df1_list = generate_upsampled_datasets(train_df1, n_datasets=10, noise_level=0.01)\n",
    "\n",
    "print(\"Generating upsampled datasets for Dataset 2...\")\n",
    "upsampled_df2_list = generate_upsampled_datasets(train_df2, n_datasets=10, noise_level=0.01)\n",
    "\n",
    "# STEP 4 — SAVE UPSAMPLED DATASETS\n",
    "\n",
    "\n",
    "# Save original training sets\n",
    "train_df1.to_csv(os.path.join(OUTPUT_DIR, \"dataset1_train_original.csv\"), index=False)\n",
    "train_df2.to_csv(os.path.join(OUTPUT_DIR, \"dataset2_train_original.csv\"), index=False)\n",
    "\n",
    "# Save augmented datasets\n",
    "for i, df_aug in enumerate(upsampled_df1_list):\n",
    "    df_aug.to_csv(os.path.join(OUTPUT_DIR, f\"dataset1_train_augmented_{i+1}.csv\"), index=False)\n",
    "\n",
    "for i, df_aug in enumerate(upsampled_df2_list):\n",
    "    df_aug.to_csv(os.path.join(OUTPUT_DIR, f\"dataset2_train_augmented_{i+1}.csv\"), index=False)\n",
    "\n",
    "print(\"All augmented datasets saved in /data folder.\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "267c8264-5f2d-4673-96e0-b89c7ae7fa0d",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}