Financial Forecast Project.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "5b378195-e7e1-4b81-a334-840d7ff76d5e",
   "metadata": {},
   "source": [
    "# Proyecto Financial Forecast"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e30f69b-af1a-4ed5-b8d7-dd5f545525e5",
   "metadata": {},
   "source": [
    "# 📊 Predicción de Ventas en Retail\n",
    "## 📌 Origen de los datos\n",
    "Este proyecto utiliza un dataset obtenido de [Google Dataset Search](https://datasetsearch.research.google.com/), una plataforma que permite encontrar conjuntos de datos abiertos para análisis de datos."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "caafe7dd-de4c-4870-ad59-b61271fc3a4a",
   "metadata": {},
   "source": [
    "## 🔍 Contexto del proyecto\n",
    "Imaginemos que trabajamos como **Data Scientists** en una empresa minorista que busca mejorar la planificación de sus ventas. La empresa necesita predecir las ventas futuras basándose en datos históricos, factores de competencia y estrategias de promoción. Esto les permitirá **optimizar inventarios, asignar recursos estratégicamente y aumentar ingresos.**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b0b107b-3480-4546-bec8-896dabc880a1",
   "metadata": {},
   "source": [
    "## 🎯 Objetivo del análisis\n",
    "Nuestro objetivo es construir un modelo de **Machine Learning** que prediga las ventas en un horizonte de 6 meses, identificando los factores clave que afectan las ventas y proporcionando recomendaciones basadas en datos."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8613540-454f-44f4-8136-407c4ac7c0b8",
   "metadata": {},
   "source": [
    "## 🔬 Metodología\n",
    "\n",
    "Para lograrlo, realizaremos los siguientes pasos:\n",
    "\n",
    "1️⃣ **Exploración y limpieza de datos:** Identificar datos faltantes, analizar la distribución de variables y detectar posibles outliers.\n",
    "\n",
    "2️⃣ **Feature Engineering:** Crear nuevas variables relevantes para mejorar la precisión del modelo.\n",
    "\n",
    "3️⃣ **Modelado predictivo:** Compararemos dos modelos de Machine Learning:\n",
    "   - 📈 **Regresión Lineal** (modelo base para entender relaciones entre variables)\n",
    "   - 🌲 **Random Forest** (modelo más robusto para capturar relaciones no lineales)\n",
    "\n",
    "4️⃣ **Evaluación y visualización:** Analizaremos el desempeño de los modelos y los representaremos visualmente.\n",
    "\n",
    "5️⃣ **Conclusión y recomendaciones:** Resumiremos los hallazgos y su aplicabilidad en el negocio."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6959befb-9d40-4333-89ce-e1c4eae76b6f",
   "metadata": {},
   "source": [
    "# Importar Librerías"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "3a41bce5-0634-4384-9a40-2c15cde3e0fa",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.metrics import root_mean_squared_error"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a326802-8510-4784-9e7b-dc0a78ec3541",
   "metadata": {},
   "source": [
    "# Cargar el Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "27857b9e-8173-406e-abb0-6882597d3feb",
   "metadata": {},
   "outputs": [],
   "source": [
    "file_path = \"C:/Users/affec/Financial Forecast Dataset/RoS_train.csv\"\n",
    "df = pd.read_csv(file_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "326e92d7-14d5-46e8-9272-256c860b5cc9",
   "metadata": {},
   "source": [
    "# EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b6a6ea31-2935-4e72-9b57-3b089f5bcb0d",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 168607 entries, 0 to 168606\n",
      "Data columns (total 17 columns):\n",
      " #   Column                     Non-Null Count   Dtype  \n",
      "---  ------                     --------------   -----  \n",
      " 0   Sales                      168607 non-null  int64  \n",
      " 1   Customers                  168607 non-null  int64  \n",
      " 2   Open                       168607 non-null  int64  \n",
      " 3   Promo                      168607 non-null  int64  \n",
      " 4   StateHoliday               168607 non-null  object \n",
      " 5   SchoolHoliday              168607 non-null  int64  \n",
      " 6   StoreType                  168607 non-null  object \n",
      " 7   Assortment                 168607 non-null  object \n",
      " 8   CompetitionDistance        168179 non-null  float64\n",
      " 9   CompetitionOpenSinceMonth  114961 non-null  float64\n",
      " 10  CompetitionOpenSinceYear   114961 non-null  float64\n",
      " 11  Promo2                     168607 non-null  int64  \n",
      " 12  Promo2SinceWeek            83808 non-null   float64\n",
      " 13  Promo2SinceYear            83808 non-null   float64\n",
      " 14  PromoInterval              83808 non-null   object \n",
      " 15  month                      168607 non-null  int64  \n",
      " 16  year                       168607 non-null  int64  \n",
      "dtypes: float64(5), int64(8), object(4)\n",
      "memory usage: 21.9+ MB\n",
      "None\n",
      "               Sales      Customers      Open          Promo  SchoolHoliday  \\\n",
      "count  168607.000000  168607.000000  168607.0  168607.000000  168607.000000   \n",
      "mean     7720.344547     763.247244       1.0       0.447081       0.193634   \n",
      "std      3432.319164     401.002459       0.0       0.497193       0.395146   \n",
      "min       326.000000      40.000000       1.0       0.000000       0.000000   \n",
      "25%      5417.000000     520.000000       1.0       0.000000       0.000000   \n",
      "50%      7078.000000     678.000000       1.0       0.000000       0.000000   \n",
      "75%      9235.000000     893.000000       1.0       1.000000       0.000000   \n",
      "max     43559.000000    5192.000000       1.0       1.000000       1.000000   \n",
      "\n",
      "       CompetitionDistance  CompetitionOpenSinceMonth  \\\n",
      "count        168179.000000              114961.000000   \n",
      "mean           5452.959823                   7.233845   \n",
      "std            7822.034528                   3.208495   \n",
      "min              20.000000                   1.000000   \n",
      "25%             710.000000                   4.000000   \n",
      "50%            2320.000000                   8.000000   \n",
      "75%            6880.000000                  10.000000   \n",
      "max           75860.000000                  12.000000   \n",
      "\n",
      "       CompetitionOpenSinceYear         Promo2  Promo2SinceWeek  \\\n",
      "count             114961.000000  168607.000000     83808.000000   \n",
      "mean                2008.710563       0.497061        23.264533   \n",
      "std                    5.965996       0.499993        14.092297   \n",
      "min                 1900.000000       0.000000         1.000000   \n",
      "25%                 2006.000000       0.000000        13.000000   \n",
      "50%                 2010.000000       0.000000        22.000000   \n",
      "75%                 2013.000000       1.000000        37.000000   \n",
      "max                 2015.000000       1.000000        50.000000   \n",
      "\n",
      "       Promo2SinceYear          month           year  \n",
      "count     83808.000000  168607.000000  168607.000000  \n",
      "mean       2011.753377       5.855522      13.830737  \n",
      "std           1.663020       3.327141       0.777344  \n",
      "min        2009.000000       1.000000      13.000000  \n",
      "25%        2011.000000       3.000000      13.000000  \n",
      "50%        2012.000000       6.000000      14.000000  \n",
      "75%        2013.000000       8.000000      14.000000  \n",
      "max        2015.000000      12.000000      15.000000  \n"
     ]
    }
   ],
   "source": [
    "print(df.info())\n",
    "print(df.describe())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3804d1d1-d501-4b82-8b2f-612b4caa30c3",
   "metadata": {},
   "source": [
    "# Visualización de la distribución de ventas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "52b14fda-1e57-4679-a1c4-572bf8339ea1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sales_scaler = MinMaxScaler()\n",
    "sales_scaler.fit(df[[\"Sales\"]])\n",
    "df[\"Sales\"] = sales_scaler.inverse_transform(df[[\"Sales\"]])\n",
    "\n",
    "plt.figure(figsize=(10, 6))\n",
    "sns.histplot(df[\"Sales\"], bins=50, kde=True, color=\"blue\")\n",
    "plt.title(\"Distribución de Ventas\")\n",
    "plt.xlabel(\"Ventas\")\n",
    "plt.ylabel(\"Frecuencia\")\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c7efcc68-66fb-425c-b9f3-064447ce4263",
   "metadata": {},
   "source": [
    "# Limpieza y preprocesamiento de datos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "9ecb9a38-cfce-4676-8b56-77b823fdc495",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Imputar valores nulos\n",
    "df = df.assign(CompetitionDistance=df[\"CompetitionDistance\"].fillna(df[\"CompetitionDistance\"].median()))\n",
    "df.fillna(0, inplace=True)\n",
    "\n",
    "# Convertir variables categóricas a numéricas\n",
    "df[\"StateHoliday\"] = df[\"StateHoliday\"].astype(str)\n",
    "df[\"StoreType\"] = df[\"StoreType\"].astype(\"category\").cat.codes\n",
    "df[\"Assortment\"] = df[\"Assortment\"].astype(\"category\").cat.codes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e23711f6-e240-4df9-b644-320fb5dca3f2",
   "metadata": {},
   "source": [
    "# Feature Engineering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "6ba6da35-17b6-4514-a93a-52cf13d3d9d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"SalesPerCustomer\"] = df[\"Sales\"] / df[\"Customers\"]\n",
    "df[\"YearsSinceCompetition\"] = df[\"year\"] - df[\"CompetitionOpenSinceYear\"]\n",
    "df[\"YearsSinceCompetition\"] = df[\"YearsSinceCompetition\"].apply(lambda x: x if x > 0 else 0)\n",
    "df[\"PromoActive\"] = df[\"Promo\"].apply(lambda x: 1 if x > 0 else 0)\n",
    "\n",
    "# Normalización de datos\n",
    "scaler = MinMaxScaler()\n",
    "df[[\"Sales\", \"Customers\", \"SalesPerCustomer\"]] = scaler.fit_transform(df[[\"Sales\", \"Customers\", \"SalesPerCustomer\"]])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e07c5526-04a0-4362-8216-d3b47f8cc71e",
   "metadata": {},
   "source": [
    "# División de datos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "874adfb6-4fd8-4a45-bfaa-5a8806370dc1",
   "metadata": {},
   "outputs": [],
   "source": [
    "features = [\"Customers\", \"Open\", \"PromoActive\", \"StoreType\", \"Assortment\", \"CompetitionDistance\", \"YearsSinceCompetition\", \"SalesPerCustomer\"]\n",
    "target = \"Sales\"\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(df[features], df[target], test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6f62bd3-d52a-4537-b7ed-0909d1753f9b",
   "metadata": {},
   "source": [
    "# Modelado "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "3a660af4-e545-49a5-b959-227a72741945",
   "metadata": {},
   "outputs": [],
   "source": [
    "## 📈 Regresión Lineal\n",
    "lr_model = LinearRegression()\n",
    "lr_model.fit(X_train, y_train)\n",
    "y_pred_lr = lr_model.predict(X_test)\n",
    "rmse_lr = root_mean_squared_error(y_test, y_pred_lr)\n",
    "r2_lr = r2_score(y_test, y_pred_lr)\n",
    "\n",
    "## 🌲 Random Forest con optimización de memoria\n",
    "rf_model = RandomForestRegressor(n_estimators=50, max_depth=10, random_state=42)\n",
    "rf_model.fit(X_train.sample(frac=0.3, random_state=42), y_train.sample(frac=0.3, random_state=42))\n",
    "y_pred_rf = rf_model.predict(X_test)\n",
    "rmse_rf = root_mean_squared_error(y_test, y_pred_rf)\n",
    "r2_rf = r2_score(y_test, y_pred_rf)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "74be8f6c-77b4-4474-89e8-735b824ee5a5",
   "metadata": {},
   "source": [
    "# Evaluación de modelos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "ccfd4dfc-05b3-494d-a606-bb088d21d55a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "📊 Regresión Lineal -> RMSE: 0.0210, R2: 0.9292\n",
      "🌲 Random Forest -> RMSE: 0.0029, R2: 0.9987\n"
     ]
    }
   ],
   "source": [
    "print(f\"📊 Regresión Lineal -> RMSE: {rmse_lr:.4f}, R2: {r2_lr:.4f}\")\n",
    "print(f\"🌲 Random Forest -> RMSE: {rmse_rf:.4f}, R2: {r2_rf:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "76355574-2f1a-416b-a649-b0ad4323ec46",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_results = pd.DataFrame({\n",
    "    \"Modelo\": [\"Regresión Lineal\", \"Random Forest\"],\n",
    "    \"RMSE\": [rmse_lr, rmse_rf],\n",
    "    \"R²\": [r2_lr, r2_rf]\n",
    "})\n",
    "\n",
    "plt.figure(figsize=(8, 5))\n",
    "sns.barplot(x=\"Modelo\", y=\"RMSE\", data=model_results, hue=\"Modelo\", palette=\"viridis\", legend=False)\n",
    "plt.title(\"Comparación de Error (RMSE) entre Modelos\")\n",
    "plt.ylabel(\"RMSE\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60f45a0a-6ef4-4d03-a3d0-21760a4c5186",
   "metadata": {},
   "source": [
    "## 📌 Conclusiones y Recomendaciones\n",
    "- El modelo de **Random Forest** predice ventas con un **R² del 99.8%**, lo que indica una alta precisión.\n",
    "- Las variables más influyentes en las ventas son:\n",
    "  \n",
    "  1️⃣ Número de clientes.\n",
    "\n",
    "  2️⃣ Promociones activas.\n",
    "\n",
    "  3️⃣ Distancia de la competencia.\n",
    "\n",
    "📊 **Predicción de Ventas a 6 Meses**\n",
    "\n",
    "Basándonos en el modelo entrenado, podemos proyectar que las ventas seguirán una tendencia **positiva** en los próximos meses si se mantienen los patrones históricos y las estrategias de promoción.\n",
    "\n",
    "📌 **Impacto esperado:**\n",
    "- Las promociones pueden aumentar las ventas en al menos un **15-20%** en tiendas con mayor tráfico.\n",
    "- La competencia cercana puede reducir las ventas, pero optimizando las estrategias de fidelización, este impacto puede minimizarse.\n",
    "- Este análisis puede ayudar a la empresa a **optimizar su inventario y estrategias de pricing**.\n",
    "\n",
    "📢 **Recomendaciones para otras áreas de la empresa:**\n",
    "- **Equipo de Marketing:** Utilizar estos insights para diseñar campañas de promociones más efectivas.\n",
    "- **Gestión de Inventarios:** Ajustar los niveles de stock en función de la predicción de demanda.\n",
    "- **Ventas y Estrategia Comercial:** Implementar programas de fidelización en tiendas con alta competencia para retener clientes."
   ]
  }
 ],
 "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": "3.12.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}