Notebooks for module on intermittent time series forecasting

marcopeix · Sep 30, 2023 · 044bf28 · 044bf28
1 parent a741680
commit 044bf28
Show file tree

Hide file tree

Showing 10 changed files with 2,922 additions and 0 deletions.
diff --git a/26_croston.ipynb b/26_croston.ipynb
diff --git a/26_croston_starter.ipynb b/26_croston_starter.ipynb
@@ -0,0 +1,88 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "81a34b98",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "import warnings\n",
+    "warnings.filterwarnings('ignore')\n",
+    "\n",
+    "%matplotlib inline"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "e1cf585c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "plt.rcParams[\"figure.figsize\"] = (9,6)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "97dfdc88",
+   "metadata": {},
+   "source": [
+    "# Croston's method"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "67466e30",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df = pd.read_csv('data/intermittent_time_series.csv')\n",
+    "\n",
+    "df.head()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "2389581d",
+   "metadata": {},
+   "source": [
+    "## Optimized Croston's Method "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "6c116ec6",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "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": "3.10.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
diff --git a/27_adida_imapa.ipynb b/27_adida_imapa.ipynb
diff --git a/27_adida_imapa_starter.ipynb b/27_adida_imapa_starter.ipynb
@@ -0,0 +1,196 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "6027e279",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "from statsforecast import StatsForecast\n",
+    "from statsforecast.models import CrostonOptimized\n",
+    "\n",
+    "import warnings\n",
+    "warnings.filterwarnings('ignore')\n",
+    "\n",
+    "%matplotlib inline"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "89c57f05",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "plt.rcParams[\"figure.figsize\"] = (9,6)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "224ffed1",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df = pd.read_csv('data/intermittent_time_series.csv')\n",
+    "\n",
+    "df.head()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "df4c6ea3",
+   "metadata": {},
+   "source": [
+    "## ADIDA "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d8be5a99",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Model with ADIDA and Croston"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "768d18c8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig, ax = plt.subplots(figsize=(10,8))\n",
+    "\n",
+    "ax.bar(df.index, df['y'], color='lightgray')\n",
+    "ax.plot(cv_df.index, cv_df['CrostonOptimized'], ls='--', label='Croston (optimized)')\n",
+    "ax.plot(cv_df.index, cv_df['ADIDA'], ls=':', label='ADIDA')\n",
+    "ax.set_ylabel('Value')\n",
+    "ax.set_xlabel('Time steps')\n",
+    "ax.legend(loc='best')\n",
+    "plt.xlim(40, 100)\n",
+    "\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "feed0476",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Calculate the MAE\n",
+    "from sklearn.metrics import mean_absolute_error\n",
+    "\n",
+    "\n",
+    "print(f'MAE Croston: {mae_croston}')\n",
+    "print(f'MAE ADIDA: {mae_adida}')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "f28706aa",
+   "metadata": {},
+   "source": [
+    "## IMAPA "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8ed3259f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Model with IMAPA and ADIDA\n",
+    "\n",
+    "sf = StatsForecast(\n",
+    "    df=df,\n",
+    "    models=models,\n",
+    "    freq='H',\n",
+    "    n_jobs=-1\n",
+    ")\n",
+    "\n",
+    "cv_df = sf.cross_validation(\n",
+    "    df=df,\n",
+    "    h=1,\n",
+    "    step_size=1,\n",
+    "    n_windows=50\n",
+    ")\n",
+    "\n",
+    "cv_df.index = np.arange(50, 100, 1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "0e7710e0",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig, ax = plt.subplots(figsize=(10,8))\n",
+    "\n",
+    "ax.bar(df.index, df['y'], color='lightgray')\n",
+    "ax.plot(cv_df.index, cv_df['IMAPA'], ls='--', label='IMAPA')\n",
+    "ax.plot(cv_df.index, cv_df['ADIDA'], ls=':', label='ADIDA')\n",
+    "ax.set_ylabel('Value')\n",
+    "ax.set_xlabel('Time steps')\n",
+    "ax.legend(loc='best')\n",
+    "plt.xlim(40, 100)\n",
+    "\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "87d6cbe5",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "mae_imapa = mean_absolute_error(cv_df['y'], cv_df['IMAPA'])\n",
+    "mae_adida = mean_absolute_error(cv_df['y'], cv_df['ADIDA'])\n",
+    "\n",
+    "print(f'MAE Croston: {mae_imapa}')\n",
+    "print(f'MAE ADIDA: {mae_adida}')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "a98074fa",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "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": "3.10.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}