Spaces:
Sleeping
Sleeping
File size: 16,331 Bytes
6011734 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 | {
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"colab": {
"provenance": []
},
"kernelspec": {
"name": "python3",
"display_name": "Python 3"
},
"language_info": {
"name": "python"
}
},
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"id": "2ERyVGhbyopK"
},
"outputs": [],
"source": [
"# Role: Data Analyst\n",
"# Pipeline:\n",
"# CLEAN > ENCODE > SPLIT 80-20 > RANDOM FOREST CLASSIFICATION (satisfaction)\n",
"# > ARIMA REVENUE FORECAST > FEATURE IMPORTANCE > EVALUATION\n",
"# =============================================================================\n",
"\n",
"import pandas as pd\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"import matplotlib.gridspec as gridspec\n",
"import seaborn as sns\n",
"import warnings\n",
"warnings.filterwarnings(\"ignore\")\n",
"\n",
"from sklearn.ensemble import RandomForestClassifier\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.preprocessing import LabelEncoder\n",
"from sklearn.metrics import (classification_report,\n",
" ConfusionMatrixDisplay,\n",
" accuracy_score)\n",
"from statsmodels.tsa.arima.model import ARIMA\n",
"\n",
"PALETTE = [\"#2E4057\", \"#048A81\", \"#54C6EB\", \"#EFD28D\", \"#C84B31\"]\n",
"sns.set_theme(style=\"whitegrid\", palette=PALETTE)\n",
"plt.rcParams.update({\"figure.dpi\": 130, \"axes.titlesize\": 13,\n",
" \"axes.labelsize\": 11})"
]
},
{
"cell_type": "code",
"source": [
"# SECTION 1 \u2013 LOAD DATA FROM NOTEBOOK 1\n",
"# =============================================================================\n",
"\n",
"ride_df = pd.read_csv(\"ride_data_clean.csv\")\n",
"review_df = pd.read_csv(\"review_data_clean.csv\")\n",
"merged_df = pd.read_csv(\"merged_summary.csv\")\n",
"\n",
"print(f\"Rides: {ride_df.shape} | Reviews: {review_df.shape} | Merged: {merged_df.shape}\")"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "63oe81VNzi8_",
"outputId": "245fcee0-e3b0-41ed-9fb3-3ec7c4fd6eba"
},
"execution_count": 3,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"Rides: (1000, 12) | Reviews: (1500, 7) | Merged: (16, 11)\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"# SECTION 2 \u2013 CLASSIFICATION: PREDICT USER SATISFACTION (HIGH vs LOW)\n",
"# Dependent variable : SatisfactionLabel (High = rating \u2265 4, Low otherwise)\n",
"# Independent variables: final_price_eur, distance_km, duration_min,\n",
"# discount_pct, cancelled, ride_type, time_slot\n",
"# =============================================================================\n",
"\n",
"# \u2500\u2500 2a. Build classification dataframe \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n",
"clf_df = ride_df[[\n",
" \"final_price_eur\", \"distance_km\", \"duration_min\",\n",
" \"discount_pct\", \"cancelled\", \"ride_type\", \"time_slot\", \"rating\"\n",
"]].copy()\n",
"\n",
"clf_df[\"SatisfactionLabel\"] = (clf_df[\"rating\"] >= 4).astype(int) # 1=High, 0=Low\n",
"clf_df.drop(columns=\"rating\", inplace=True)\n",
"\n",
"# \u2500\u2500 2b. Encode categoricals \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n",
"le_rt = LabelEncoder()\n",
"le_ts = LabelEncoder()\n",
"clf_df[\"ride_type_enc\"] = le_rt.fit_transform(clf_df[\"ride_type\"])\n",
"clf_df[\"time_slot_enc\"] = le_ts.fit_transform(clf_df[\"time_slot\"])\n",
"clf_df.drop(columns=[\"ride_type\", \"time_slot\"], inplace=True)\n",
"\n",
"X = clf_df.drop(columns=\"SatisfactionLabel\")\n",
"y = clf_df[\"SatisfactionLabel\"]\n",
"\n",
"# \u2500\u2500 2c. Train / test split 80-20 \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n",
"X_train, X_test, y_train, y_test = train_test_split(\n",
" X, y, test_size=0.20, random_state=42, stratify=y)\n",
"\n",
"print(f\"\\nTrain size: {len(X_train)} | Test size: {len(X_test)}\")\n",
"\n",
"# \u2500\u2500 2d. Random Forest \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n",
"rf = RandomForestClassifier(n_estimators=200, max_depth=8,\n",
" random_state=42, class_weight=\"balanced\")\n",
"rf.fit(X_train, y_train)\n",
"y_pred = rf.predict(X_test)\n",
"\n",
"print(f\"\\nClassification Accuracy: {accuracy_score(y_test, y_pred):.4f}\")\n",
"print(\"\\nClassification Report:\")\n",
"print(classification_report(y_test, y_pred,\n",
" target_names=[\"Low Satisfaction\", \"High Satisfaction\"]))\n"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "nv-OXM0nzywU",
"outputId": "335d1c9d-6d2b-4878-9aed-2e3e05e11905"
},
"execution_count": 4,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"\n",
"Train size: 800 | Test size: 200\n",
"\n",
"Classification Accuracy: 0.6450\n",
"\n",
"Classification Report:\n",
" precision recall f1-score support\n",
"\n",
" Low Satisfaction 0.22 0.16 0.18 50\n",
"High Satisfaction 0.74 0.81 0.77 150\n",
"\n",
" accuracy 0.65 200\n",
" macro avg 0.48 0.48 0.48 200\n",
" weighted avg 0.61 0.65 0.63 200\n",
"\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"# SECTION 3 \u2013 ARIMA REVENUE FORECAST\n",
"# Aggregate weekly total revenue \u2192 forecast next 12 weeks for 3 sample cities\n",
"# =============================================================================\n",
"\n",
"# Generate a synthetic weekly revenue time series per city (realistic trend + noise)\n",
"np.random.seed(7)\n",
"weeks = pd.date_range(\"2022-01-03\", periods=104, freq=\"W\") # 2 years weekly\n",
"cities_sel = [\"Paris\", \"Berlin\", \"Madrid\"]\n",
"\n",
"city_rev = {}\n",
"for c in cities_sel:\n",
" trend = np.linspace(80_000, 130_000, 104)\n",
" season = 8_000 * np.sin(np.linspace(0, 4 * np.pi, 104))\n",
" noise = np.random.normal(0, 5_000, 104)\n",
" city_rev[c] = pd.Series(trend + season + noise, index=weeks)\n",
"\n",
"FORECAST_STEPS = 12"
],
"metadata": {
"id": "2tsC_F9oz4JO"
},
"execution_count": 5,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# SECTION 4 \u2013 VISUALIZATIONS (5 charts)\n",
"# =============================================================================\n",
"\n",
"fig = plt.figure(figsize=(20, 24))\n",
"fig.suptitle(\"Urban Mobility \u2013 Predictive Analytics & Revenue Forecasting\",\n",
" fontsize=17, fontweight=\"bold\", y=0.99)\n",
"gs = gridspec.GridSpec(3, 2, figure=fig, hspace=0.50, wspace=0.35)\n",
"\n",
"# \u2500\u2500 Chart 1: Feature importance \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n",
"ax1 = fig.add_subplot(gs[0, 0])\n",
"feat_imp = pd.Series(rf.feature_importances_, index=X.columns).sort_values()\n",
"feat_imp.index = [\"Discount %\", \"Cancelled\", \"Ride Type\",\n",
" \"Time Slot\", \"Duration (min)\", \"Distance (km)\", \"Final Price (\u20ac)\"]\n",
"feat_imp.sort_values().plot(kind=\"barh\", ax=ax1, color=PALETTE[1])\n",
"ax1.set_title(\"Random Forest \u2013 Feature Importances\\n(Satisfaction Classification)\")\n",
"ax1.set_xlabel(\"Importance Score\")\n",
"\n",
"# \u2500\u2500 Chart 2: Confusion matrix \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n",
"ax2 = fig.add_subplot(gs[0, 1])\n",
"ConfusionMatrixDisplay.from_predictions(\n",
" y_test, y_pred,\n",
" display_labels=[\"Low\", \"High\"],\n",
" colorbar=False, cmap=\"Blues\", ax=ax2)\n",
"ax2.set_title(\"Confusion Matrix\\n(Satisfaction: Low vs High)\")\n",
"\n",
"# \u2500\u2500 Charts 3-5: ARIMA forecasts per city \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n",
"arima_positions = [(1, 0), (1, 1), (2, 0)]\n",
"arima_models = {}\n",
"\n",
"for idx, city in enumerate(cities_sel):\n",
" row, col = arima_positions[idx]\n",
" ax = fig.add_subplot(gs[row, col])\n",
" series = city_rev[city]\n",
"\n",
" model = ARIMA(series, order=(2, 1, 2))\n",
" result = model.fit()\n",
" arima_models[city] = result\n",
"\n",
" forecast = result.get_forecast(steps=FORECAST_STEPS)\n",
" forecast_df = forecast.summary_frame(alpha=0.10)\n",
" future_idx = pd.date_range(series.index[-1] + pd.Timedelta(weeks=1),\n",
" periods=FORECAST_STEPS, freq=\"W\")\n",
" forecast_df.index = future_idx\n",
"\n",
" ax.plot(series, color=PALETTE[0], linewidth=1.2, label=\"Historical\")\n",
" ax.plot(forecast_df[\"mean\"], color=PALETTE[2],\n",
" linewidth=2, linestyle=\"--\", label=\"Forecast\")\n",
" ax.fill_between(forecast_df.index,\n",
" forecast_df[\"mean_ci_lower\"],\n",
" forecast_df[\"mean_ci_upper\"],\n",
" alpha=0.25, color=PALETTE[2], label=\"90% CI\")\n",
" ax.set_title(f\"ARIMA Revenue Forecast \u2013 {city}\")\n",
" ax.set_ylabel(\"Weekly Revenue (\u20ac)\")\n",
" ax.set_xlabel(\"\")\n",
" ax.legend(fontsize=8)\n",
" ax.yaxis.set_major_formatter(\n",
" plt.FuncFormatter(lambda v, _: f\"\u20ac{v/1000:.0f}k\"))\n",
"\n",
"# \u2500\u2500 Chart 6 (last cell): Price sensitivity \u2013 avg rating by price bucket \u2500\u2500\u2500\u2500\u2500\u2500\u2500\n",
"ax6 = fig.add_subplot(gs[2, 1])\n",
"ride_df[\"price_bucket\"] = pd.cut(ride_df[\"final_price_eur\"],\n",
" bins=[0, 2, 3.5, 5, 6.5, 10],\n",
" labels=[\"<2\", \"2\u20133.5\", \"3.5\u20135\", \"5\u20136.5\", \">6.5\"])\n",
"price_sens = ride_df.groupby(\"price_bucket\", observed=True)[\"rating\"].mean()\n",
"ax6.bar(price_sens.index, price_sens.values, color=PALETTE[3], edgecolor=\"white\")\n",
"ax6.set_title(\"Price Sensitivity \u2013 Avg. Rating by Price Bucket\")\n",
"ax6.set_xlabel(\"Final Price (\u20ac)\")\n",
"ax6.set_ylabel(\"Avg. Rating\")\n",
"ax6.set_ylim(3, 5)\n",
"for p, v in zip(price_sens.index, price_sens.values):\n",
" ax6.text(p, v + 0.02, f\"{v:.2f}\u2605\", ha=\"center\", fontsize=9)\n",
"\n",
"plt.savefig(\"notebook2_models_output.png\", bbox_inches=\"tight\")\n",
"plt.close()\n",
"print(\"\\n\u2705 Model output chart saved \u2192 notebook2_models_output.png\")"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "o7Y2B81jz89H",
"outputId": "35a76ae9-1ac1-4de8-bfba-6c349af72d11"
},
"execution_count": 7,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"\n",
"\u2705 Model output chart saved \u2192 notebook2_models_output.png\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"# SECTION 5 \u2013 SAVE FORECAST TABLE\n",
"# =============================================================================\n",
"all_forecasts = []\n",
"for city in cities_sel:\n",
" fc = arima_models[city].get_forecast(steps=FORECAST_STEPS).summary_frame(alpha=0.10)\n",
" fc.index = pd.date_range(city_rev[city].index[-1] + pd.Timedelta(weeks=1),\n",
" periods=FORECAST_STEPS, freq=\"W\")\n",
" fc[\"city\"] = city\n",
" all_forecasts.append(fc[[\"city\", \"mean\", \"mean_ci_lower\", \"mean_ci_upper\"]])\n",
"\n",
"forecast_table = pd.concat(all_forecasts)\n",
"forecast_table.columns = [\"city\", \"forecast_revenue\", \"ci_lower_90\", \"ci_upper_90\"]\n",
"forecast_table.to_csv(\"arima_forecast_table.csv\")\n",
"print(\"\u2705 Forecast table saved \u2192 arima_forecast_table.csv\")\n",
"print(forecast_table.head(6).round(0).to_string())\n",
"\n",
"print(\"\\n\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\")\n",
"print(\" NOTEBOOK 2 COMPLETE\")\n",
"print(\"\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\")"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "KY_odiwF0h7r",
"outputId": "7e77d938-59cc-4241-e4b9-37dc7cb6eaf5"
},
"execution_count": 9,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"\u2705 Forecast table saved \u2192 arima_forecast_table.csv\n",
" city forecast_revenue ci_lower_90 ci_upper_90\n",
"2024-01-07 Paris 124853.0 115042.0 134665.0\n",
"2024-01-14 Paris 124946.0 112850.0 137043.0\n",
"2024-01-21 Paris 124678.0 110776.0 138581.0\n",
"2024-01-28 Paris 124855.0 109164.0 140546.0\n",
"2024-02-04 Paris 124759.0 107576.0 141942.0\n",
"2024-02-11 Paris 124808.0 106197.0 143420.0\n",
"\n",
"\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\n",
" NOTEBOOK 2 COMPLETE\n",
"\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\n"
]
}
]
}
]
} |