[Doc] Added HierE2E Favorita baseline (#217)

* First iteration of HierE2E execution * Corrected HierE2E pipeline * Added relMSE and MSSE evaluation * Added relMSE and MSSE evaluation * Added note on available methods and predict_insample methods * Installed hooks and cleaned notebook * Fixed missing favicon * Fixed broken Getting Started quarto link * Favorita's HierE2E, implied composite likelihood approach * Favorita's HierE2E, implied composite likelihood approach * Favorita's HierE2E, implied composite likelihood approach * Favorita's HierE2E, implied composite likelihood approach --------- Co-authored-by: Cristian <cristiani.challu@gmail.com>
Nixtla · Jun 26, 2023 · e070e82 · e070e82
1 parent 1419945
commit e070e82
Show file tree

Hide file tree

Showing 2 changed files with 895 additions and 25 deletions.
diff --git a/experiments/hierarchical_baselines/nbs/run_favorita_baselines.ipynb b/experiments/hierarchical_baselines/nbs/run_favorita_baselines.ipynb
@@ -20,16 +20,7 @@
    "cell_type": "code",
    "execution_count": 1,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "/Users/kin/anaconda3/envs/hierarchical_baselines/lib/python3.10/site-packages/statsforecast/core.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
-      "  from tqdm.autonotebook import tqdm\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "import os\n",
     "import argparse\n",
@@ -41,21 +32,30 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [],
    "source": [
     "# %%capture\n",
     "# !pip install statsforecast\n",
     "# !pip install hierarchicalforecast\n",
-    "# !pip install git+https://github.com/Nixtla/datasetsforecast@feat/favorita_dataset.git"
+    "# !pip install git+https://github.com/Nixtla/datasetsforecast.git@feat/favorita_dataset"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 3,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/Users/kin/anaconda3/envs/hierarchical_baselines/lib/python3.10/site-packages/statsforecast/core.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
+      "  from tqdm.autonotebook import tqdm\n"
+     ]
+    }
+   ],
    "source": [
     "from statsforecast.core import StatsForecast\n",
     "from statsforecast.models import AutoARIMA, Naive\n",
@@ -77,7 +77,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -166,16 +166,50 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 5,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 930M/930M [00:21<00:00, 43.4MiB/s] \n",
+      "INFO:datasetsforecast.utils:Successfully downloaded favorita-grocery-sales-forecasting2.zip, 930159981, bytes.\n",
+      "INFO:datasetsforecast.utils:Decompressing zip file...\n",
+      "INFO:datasetsforecast.utils:Successfully decompressed data/favorita/favorita-grocery-sales-forecasting2.zip\n",
+      "INFO:datasetsforecast.utils:Decompressing zip file...\n",
+      "INFO:datasetsforecast.utils:Successfully decompressed data/favorita/holidays_events.csv.zip\n",
+      "INFO:datasetsforecast.utils:Decompressing zip file...\n",
+      "INFO:datasetsforecast.utils:Successfully decompressed data/favorita/items.csv.zip\n",
+      "INFO:datasetsforecast.utils:Decompressing zip file...\n",
+      "INFO:datasetsforecast.utils:Successfully decompressed data/favorita/oil.csv.zip\n",
+      "INFO:datasetsforecast.utils:Decompressing zip file...\n",
+      "INFO:datasetsforecast.utils:Successfully decompressed data/favorita/sample_submission.csv.zip\n",
+      "INFO:datasetsforecast.utils:Decompressing zip file...\n",
+      "INFO:datasetsforecast.utils:Successfully decompressed data/favorita/stores.csv.zip\n",
+      "INFO:datasetsforecast.utils:Decompressing zip file...\n",
+      "INFO:datasetsforecast.utils:Successfully decompressed data/favorita/test.csv.zip\n",
+      "INFO:datasetsforecast.utils:Decompressing zip file...\n",
+      "INFO:datasetsforecast.utils:Successfully decompressed data/favorita/train.csv.zip\n",
+      "INFO:datasetsforecast.utils:Decompressing zip file...\n",
+      "INFO:datasetsforecast.utils:Successfully decompressed data/favorita/transactions.csv.zip\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "saved train.csv to train.feather for fast access\n"
+     ]
+    }
+   ],
    "source": [
     "data = FavoritaHierarchicalDataset.load_item_data(item_id=1916577, directory = './data/favorita', dataset='Favorita200')"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -280,7 +314,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [
     {
@@ -293,16 +327,16 @@
       "\n",
       "Favorita200 item_id=112830, h=34 n_series=93, n_bottom=54 \n",
       "test ds=[(Timestamp('2017-07-13 00:00:00'), Timestamp('2017-08-15 00:00:00'))] \n",
-      "Code block 'Read and Parse data   ' took:\t0.56446 seconds\n",
-      "Code block 'Fit/Predict Model\t  ' took:\t0.13728 seconds\n",
-      "Code block 'Reconcile Predictions ' took:\t35.96450 seconds\n",
+      "Code block 'Read and Parse data   ' took:\t0.60365 seconds\n",
+      "Code block 'Fit/Predict Model\t  ' took:\t30.04792 seconds\n",
+      "Code block 'Reconcile Predictions ' took:\t38.54011 seconds\n",
       "\n",
       "\n",
       "Favorita200 item_id=1916577, h=34 n_series=93, n_bottom=54 \n",
       "test ds=[(Timestamp('2017-07-13 00:00:00'), Timestamp('2017-08-15 00:00:00'))] \n",
-      "Code block 'Read and Parse data   ' took:\t0.58135 seconds\n",
-      "Code block 'Fit/Predict Model\t  ' took:\t0.12099 seconds\n",
-      "Code block 'Reconcile Predictions ' took:\t35.64888 seconds\n",
+      "Code block 'Read and Parse data   ' took:\t0.61150 seconds\n",
+      "Code block 'Fit/Predict Model\t  ' took:\t29.67945 seconds\n",
+      "Code block 'Reconcile Predictions ' took:\t38.78811 seconds\n",
       "\n",
       "\n"
      ]