{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Forecasting Walmart sales with Treeffuser\n", "\n", "In this tutorial we show how to use Treeffuser to model and forecast Walmart sales using the M5 forecasting dataset from Kaggle." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Getting started\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To get started, we first install `treeffuser` and import the relevant libraries." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "%%capture\n", "!pip install treeffuser\n", "\n", "import matplotlib.pyplot as plt\n", "import numpy as np\n", "import pandas as pd\n", "\n", "from pathlib import Path\n", "\n", "from tqdm import tqdm\n", "from treeffuser import Treeffuser\n", "\n", "# load autoreload extension\n", "%load_ext autoreload\n", "%autoreload 2" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Next, create a Kaggle account and download the data from https://www.kaggle.com/competitions/m5-forecasting-accuracy/data.\n", "\n", "If you're running this notebook in Colab, manually upload the necessary files (`calendar.csv`, `sales_train_validation.csv`, `sell_prices.csv`) to Colab by clicking the `Files` tab on the left sidebar and selecting `Upload`. Move the files into a new folder named `m5`. Once uploaded, the notebook will be able to read and process the data.\n", "\n", "If you're running this on your local machine, you can also use Kaggle's [command-line tool](https://www.kaggle.com/docs/api) and run the following from the command line:\n", "\n", "```bash\n", "cd ./m5 # path to folder where you want to save the data\n", "kaggle competitions download -c m5-forecasting-accuracy\n", "```\n", "\n", "Use your favorite tool to unzip the archive. In Linux/macOS,\n", "\n", "```bash\n", "unzip m5-forecasting-accuracy.zip\n", "```\n", "\n", "We'll be using the following files: `calendar.csv`, `sales_train_validation.csv`, and `sell_prices.csv`.\n", "\n", "\n", "\n", "\n", "" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Load the data" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "data_path = Path(\"./m5\") # change with path where you extracted the data archive\n", "\n", "calendar_df = pd.read_csv(data_path / \"calendar.csv\")\n", "sales_train_df = pd.read_csv(data_path / \"sales_train_validation.csv\")\n", "sell_prices_df = pd.read_csv(data_path / \"sell_prices.csv\")\n", "\n", "# add explicit columns for the day, month, year for ease of processing\n", "calendar_df[\"date\"] = pd.to_datetime(calendar_df[\"date\"])\n", "calendar_df[\"day\"] = calendar_df[\"date\"].dt.day\n", "calendar_df[\"month\"] = calendar_df[\"date\"].dt.month\n", "calendar_df[\"year\"] = calendar_df[\"date\"].dt.year" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## The data\n", "\n", "### Preprocessing\n", "`sell_prices_df` contains the prices of each item in each store at a given time. The `wm_yr_wk` is a unique identifier for the time." ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
store_iditem_idwm_yr_wksell_price
0CA_1HOBBIES_1_001113259.58
1CA_1HOBBIES_1_001113269.58
2CA_1HOBBIES_1_001113278.26
3CA_1HOBBIES_1_001113288.26
4CA_1HOBBIES_1_001113298.26
\n", "
" ], "text/plain": [ " store_id item_id wm_yr_wk sell_price\n", "0 CA_1 HOBBIES_1_001 11325 9.58\n", "1 CA_1 HOBBIES_1_001 11326 9.58\n", "2 CA_1 HOBBIES_1_001 11327 8.26\n", "3 CA_1 HOBBIES_1_001 11328 8.26\n", "4 CA_1 HOBBIES_1_001 11329 8.26" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sell_prices_df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`calendar_df` contains information about the dates on which the products were sold." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
datewm_yr_wkweekdaywdaymonthyeardevent_name_1event_type_1event_name_2event_type_2snap_CAsnap_TXsnap_WIday
02011-01-2911101Saturday112011d_1NaNNaNNaNNaN00029
12011-01-3011101Sunday212011d_2NaNNaNNaNNaN00030
22011-01-3111101Monday312011d_3NaNNaNNaNNaN00031
32011-02-0111101Tuesday422011d_4NaNNaNNaNNaN1101
42011-02-0211101Wednesday522011d_5NaNNaNNaNNaN1012
\n", "
" ], "text/plain": [ " date wm_yr_wk weekday wday month year d event_name_1 \\\n", "0 2011-01-29 11101 Saturday 1 1 2011 d_1 NaN \n", "1 2011-01-30 11101 Sunday 2 1 2011 d_2 NaN \n", "2 2011-01-31 11101 Monday 3 1 2011 d_3 NaN \n", "3 2011-02-01 11101 Tuesday 4 2 2011 d_4 NaN \n", "4 2011-02-02 11101 Wednesday 5 2 2011 d_5 NaN \n", "\n", " event_type_1 event_name_2 event_type_2 snap_CA snap_TX snap_WI day \n", "0 NaN NaN NaN 0 0 0 29 \n", "1 NaN NaN NaN 0 0 0 30 \n", "2 NaN NaN NaN 0 0 0 31 \n", "3 NaN NaN NaN 1 1 0 1 \n", "4 NaN NaN NaN 1 0 1 2 " ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "calendar_df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`sales_train_df` contains the number of units sold for an item in each department and store. The sales are grouped by day: for example, the `d_1907` column has the number of units sold on the 1907-th day." ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
iditem_iddept_idcat_idstore_idstate_idd_1d_2d_3d_4...d_1904d_1905d_1906d_1907d_1908d_1909d_1910d_1911d_1912d_1913
0HOBBIES_1_001_CA_1_validationHOBBIES_1_001HOBBIES_1HOBBIESCA_1CA0000...1301113011
1HOBBIES_1_002_CA_1_validationHOBBIES_1_002HOBBIES_1HOBBIESCA_1CA0000...0000010000
2HOBBIES_1_003_CA_1_validationHOBBIES_1_003HOBBIES_1HOBBIESCA_1CA0000...2121110111
3HOBBIES_1_004_CA_1_validationHOBBIES_1_004HOBBIES_1HOBBIESCA_1CA0000...1054101372
4HOBBIES_1_005_CA_1_validationHOBBIES_1_005HOBBIES_1HOBBIESCA_1CA0000...2110112224
\n", "

5 rows × 1919 columns

\n", "
" ], "text/plain": [ " id item_id dept_id cat_id store_id \\\n", "0 HOBBIES_1_001_CA_1_validation HOBBIES_1_001 HOBBIES_1 HOBBIES CA_1 \n", "1 HOBBIES_1_002_CA_1_validation HOBBIES_1_002 HOBBIES_1 HOBBIES CA_1 \n", "2 HOBBIES_1_003_CA_1_validation HOBBIES_1_003 HOBBIES_1 HOBBIES CA_1 \n", "3 HOBBIES_1_004_CA_1_validation HOBBIES_1_004 HOBBIES_1 HOBBIES CA_1 \n", "4 HOBBIES_1_005_CA_1_validation HOBBIES_1_005 HOBBIES_1 HOBBIES CA_1 \n", "\n", " state_id d_1 d_2 d_3 d_4 ... d_1904 d_1905 d_1906 d_1907 d_1908 \\\n", "0 CA 0 0 0 0 ... 1 3 0 1 1 \n", "1 CA 0 0 0 0 ... 0 0 0 0 0 \n", "2 CA 0 0 0 0 ... 2 1 2 1 1 \n", "3 CA 0 0 0 0 ... 1 0 5 4 1 \n", "4 CA 0 0 0 0 ... 2 1 1 0 1 \n", "\n", " d_1909 d_1910 d_1911 d_1912 d_1913 \n", "0 1 3 0 1 1 \n", "1 1 0 0 0 0 \n", "2 1 0 1 1 1 \n", "3 0 1 3 7 2 \n", "4 1 2 2 2 4 \n", "\n", "[5 rows x 1919 columns]" ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sales_train_df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To align the sales data with the other DataFrames, we convert `sales_train_df` to a long format. We collapse the daily sales columns `d_{i}` into a single `sales` column, with an additional `day` column indicating the day corresponding to each sales entry." ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
item_iddept_idcat_idstore_idstate_iddsales
0HOBBIES_1_001HOBBIES_1HOBBIESCA_1CAd_10
1HOBBIES_1_001HOBBIES_1HOBBIESCA_1CAd_20
2HOBBIES_1_001HOBBIES_1HOBBIESCA_1CAd_30
3HOBBIES_1_001HOBBIES_1HOBBIESCA_1CAd_40
4HOBBIES_1_001HOBBIES_1HOBBIESCA_1CAd_50
\n", "
" ], "text/plain": [ " item_id dept_id cat_id store_id state_id d sales\n", "0 HOBBIES_1_001 HOBBIES_1 HOBBIES CA_1 CA d_1 0\n", "1 HOBBIES_1_001 HOBBIES_1 HOBBIES CA_1 CA d_2 0\n", "2 HOBBIES_1_001 HOBBIES_1 HOBBIES CA_1 CA d_3 0\n", "3 HOBBIES_1_001 HOBBIES_1 HOBBIES CA_1 CA d_4 0\n", "4 HOBBIES_1_001 HOBBIES_1 HOBBIES CA_1 CA d_5 0" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def convert_sales_data_from_wide_to_long(sales_df_wide):\n", " index_vars = [\"item_id\", \"dept_id\", \"cat_id\", \"store_id\", \"state_id\"]\n", " sales_df_long = pd.wide_to_long(\n", " sales_df_wide.iloc[:100, 1:],\n", " i=index_vars,\n", " j=\"day\",\n", " stubnames=[\"d\"],\n", " sep=\"_\",\n", " ).reset_index()\n", "\n", " sales_df_long = sales_df_long.rename(columns={\"d\": \"sales\", \"day\": \"d\"})\n", "\n", " sales_df_long[\"d\"] = \"d_\" + sales_df_long[\"d\"].astype(\n", " \"str\"\n", " ) # restore \"d_{i}\" format for day\n", " return sales_df_long\n", "\n", "\n", "sales_train_df_long = convert_sales_data_from_wide_to_long(sales_train_df)\n", "sales_train_df_long.head()" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Text(0.5, 1.0, 'Number of sales over the entire timespan')" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plt.hist(\n", " sales_train_df_long[\"sales\"],\n", " bins=np.arange(0, 10 + 1.5) - 0.5,\n", " range=[0, 10],\n", " density=True,\n", ")\n", "plt.xticks(range(10))\n", "plt.ylabel(\"Frequency of number of sales\")\n", "plt.title(\"Number of sales over the entire timespan\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Train and test sets" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The dataset comprises sales data of 100 items over 1,913 days. For simplicity, we select the data from the first 365 days and discard the rest." ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "n_items = 100\n", "n_days = 1913\n" ] } ], "source": [ "print(f\"n_items = {len(sales_train_df_long['item_id'].unique())}\")\n", "print(f\"n_days = {len(sales_train_df_long['d'].unique())}\")\n", "\n", "sales_train_df_long[\"day_number\"] = sales_train_df_long[\"d\"].str.extract(\"(\\d+)\").astype(int)\n", "data = sales_train_df_long[sales_train_df_long[\"day_number\"] <= 365].copy()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We compute the lags of the previous 30 days and merge the sales, calendar, and price data together." ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
item_iddept_idcat_idstore_idstate_iddsalesday_numbersales_lag_1sales_lag_2...yearevent_name_1event_type_1event_name_2event_type_2snap_CAsnap_TXsnap_WIdaysell_price
0HOBBIES_1_002HOBBIES_1HOBBIESCA_1CAd_14101410.00.0...2011NaNNaNNaNNaN000183.97
1HOBBIES_1_002HOBBIES_1HOBBIESCA_1CAd_14201420.00.0...2011Father's dayCulturalNaNNaN000193.97
2HOBBIES_1_002HOBBIES_1HOBBIESCA_1CAd_14301430.00.0...2011NaNNaNNaNNaN000203.97
3HOBBIES_1_002HOBBIES_1HOBBIESCA_1CAd_14411440.00.0...2011NaNNaNNaNNaN000213.97
4HOBBIES_1_002HOBBIES_1HOBBIESCA_1CAd_14501451.00.0...2011NaNNaNNaNNaN000223.97
\n", "

5 rows × 53 columns

\n", "
" ], "text/plain": [ " item_id dept_id cat_id store_id state_id d sales \\\n", "0 HOBBIES_1_002 HOBBIES_1 HOBBIES CA_1 CA d_141 0 \n", "1 HOBBIES_1_002 HOBBIES_1 HOBBIES CA_1 CA d_142 0 \n", "2 HOBBIES_1_002 HOBBIES_1 HOBBIES CA_1 CA d_143 0 \n", "3 HOBBIES_1_002 HOBBIES_1 HOBBIES CA_1 CA d_144 1 \n", "4 HOBBIES_1_002 HOBBIES_1 HOBBIES CA_1 CA d_145 0 \n", "\n", " day_number sales_lag_1 sales_lag_2 ... year event_name_1 \\\n", "0 141 0.0 0.0 ... 2011 NaN \n", "1 142 0.0 0.0 ... 2011 Father's day \n", "2 143 0.0 0.0 ... 2011 NaN \n", "3 144 0.0 0.0 ... 2011 NaN \n", "4 145 1.0 0.0 ... 2011 NaN \n", "\n", " event_type_1 event_name_2 event_type_2 snap_CA snap_TX snap_WI day \\\n", "0 NaN NaN NaN 0 0 0 18 \n", "1 Cultural NaN NaN 0 0 0 19 \n", "2 NaN NaN NaN 0 0 0 20 \n", "3 NaN NaN NaN 0 0 0 21 \n", "4 NaN NaN NaN 0 0 0 22 \n", "\n", " sell_price \n", "0 3.97 \n", "1 3.97 \n", "2 3.97 \n", "3 3.97 \n", "4 3.97 \n", "\n", "[5 rows x 53 columns]" ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "n_lags = 30\n", "\n", "# sort data before computing lags\n", "data_index_vars = [\"item_id\", \"dept_id\", \"cat_id\", \"store_id\", \"state_id\"]\n", "data.sort_values(data_index_vars + [\"day_number\"], inplace=True)\n", "\n", "for lag in range(1, n_lags + 1):\n", " data[f\"sales_lag_{lag}\"] = data.groupby(by=data_index_vars)[\"sales\"].shift(lag)\n", "\n", "data = data.merge(calendar_df).merge(sell_prices_df)\n", "\n", "data.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Treeffuser can handle **categorical columns**, but the dtype of those columns must be set to `category` in the DataFrame." ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [], "source": [ "categorical_columns = [\n", " \"item_id\",\n", " \"dept_id\",\n", " \"cat_id\",\n", " \"store_id\",\n", " \"state_id\",\n", " \"d\",\n", " \"wm_yr_wk\",\n", " \"weekday\",\n", " \"event_name_1\",\n", " \"event_type_1\",\n", " \"event_name_2\",\n", " \"event_type_2\",\n", " \"snap_CA\",\n", " \"snap_TX\",\n", " \"snap_WI\",\n", "]\n", "data[categorical_columns] = data[categorical_columns].astype(\"category\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Finally, for each item, we take the first 300 days as train data and use the remaining 65 data as test data for evaluation." ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(15216, 50)\n", "(3891, 50)\n" ] } ], "source": [ "is_train = data[\"day_number\"] <= 300\n", "\n", "y_name = \"sales\"\n", "x_names = [\n", " name for name in data.columns if name != y_name and name not in [\"day_number\", \"date\"]\n", "]\n", "\n", "X_train, y_train, dates_train = (\n", " data[is_train][x_names],\n", " data[is_train][y_name],\n", " data[is_train][\"date\"],\n", ")\n", "X_test, y_test, dates_test = (\n", " data[~is_train][x_names],\n", " data[~is_train][y_name],\n", " data[~is_train][\"date\"],\n", ")\n", "\n", "print(X_train.shape)\n", "print(X_test.shape)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Probabilistic predictions with Treeffuser\n", "\n", "We regress the sales on the following covariates." ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "item_id, dept_id, cat_id, store_id, state_id, d, sales_lag_1, sales_lag_2, sales_lag_3, sales_lag_4, sales_lag_5, sales_lag_6, sales_lag_7, sales_lag_8, sales_lag_9, sales_lag_10, sales_lag_11, sales_lag_12, sales_lag_13, sales_lag_14, sales_lag_15, sales_lag_16, sales_lag_17, sales_lag_18, sales_lag_19, sales_lag_20, sales_lag_21, sales_lag_22, sales_lag_23, sales_lag_24, sales_lag_25, sales_lag_26, sales_lag_27, sales_lag_28, sales_lag_29, sales_lag_30, wm_yr_wk, weekday, wday, month, year, event_name_1, event_type_1, event_name_2, event_type_2, snap_CA, snap_TX, snap_WI, day, sell_price\n" ] } ], "source": [ "print(\", \".join(map(str, X_train.columns)))" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[LightGBM] [Warning] Met negative value in categorical features, will convert it to NaN\n", "[LightGBM] [Warning] Met negative value in categorical features, will convert it to NaN\n", "[LightGBM] [Warning] Met negative value in categorical features, will convert it to NaN\n", "[LightGBM] [Warning] Met negative value in categorical features, will convert it to NaN\n", "[LightGBM] [Warning] Categorical features with more bins than the configured maximum bin number found.\n", "[LightGBM] [Warning] For categorical features, max_bin and max_bin_by_feature may be ignored with a large number of categories.\n" ] }, { "data": { "text/html": [ "
Treeffuser(extra_lightgbm_params={}, seed=0)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "Treeffuser(extra_lightgbm_params={}, seed=0)" ] }, "execution_count": 13, "metadata": {}, "output_type": "execute_result" } ], "source": [ "model = Treeffuser(seed=0)\n", "model.fit(X_train, y_train)" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "100%|██████████| 100/100 [05:55<00:00, 3.55s/it]\n" ] } ], "source": [ "y_test_samples = model.sample(X_test, n_samples=100, seed=0, n_steps=50, verbose=True)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Newsvendor model" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We illustrate the practical relevance of accurate probabilistic predictions with an application to inventory management, using the newsvendor model \\citep{arrow1951optimal}. \n", "\n", "Assume that every day we decide how many units $q$ of an item to buy. \n", "We buy at a cost $c$ and sell at a price $p$. \n", "However, the demand $y$ is random, introducing uncertainty in our decision. \n", "The goal is to maximize the expected profit:\n", "$$\\max_{q} p~\\mathbb{E}\\left[\\min(q, y)\\right] - c q.$$\n", "The optimal solution to the newsvendor problem is to buy $q = F^{-1}\\left( \\frac{p-c}{p} \\right)$ units, where $F^{-1}$ is the quantile function of the distribution of $y$. \n", "\n", "Using Treeffuser, we can compute the quantiles from the samples and forecast the optimal quantity of units to buy.\n", "\n", "To compute profits, we use the observed prices, assume a margin of $50\\%$ over all products, and assume the actual number of sales of an item correspond to the demand of this item. We let Treeffuser, learn the conditional distribution of the demand of each item, estimate their quantiles, and thus determine the optimal quantity to buy. \n", "\n", "We use the held-out data to compute the profit made if Treeffuser was used to forecast the demand of each item and to manage the inventory of each item." ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [], "source": [ "def newsvendor_utility(y_true, quantity_ordered, prices, stocking_cost):\n", " \"\"\"\n", " The newsvendor utility function with stock q, demand y, selling price p, stocking cost c is given by\n", " $$ U(y, q, p, c) = p * min(y, q) - c * q $$\n", " \"\"\"\n", " utility = prices * np.minimum(y_true, quantity_ordered) - stocking_cost * quantity_ordered\n", " return utility\n", "\n", "\n", "def newsvendor_optimal_quantity(y_samples, prices, stocking_cost):\n", " \"\"\"\n", " Returns the optimal quantity to order for the newsvendor problem.\n", "\n", " It is given theoeretically by:\n", " $$ q* = argmax_{q} E[U(y, q, p, c)] $$\n", " which has a closed form solution,\n", " $$ q* = F^{-1}( (p - c) / p) $$\n", " where F is the CDF of the demand distribution\n", " \"\"\"\n", " # compute the target quantiles (p - c) / p\n", " target_quantiles = (prices - stocking_cost) / prices\n", " target_quantiles = np.maximum(target_quantiles, 0.0)\n", "\n", " # compute the empirical quantities corresponding to the target quantiles\n", " res = []\n", " for i in range(y_samples.shape[1]):\n", " optimal_quantities = np.quantile(y_samples[:, i], target_quantiles[i])\n", " res.append(optimal_quantities)\n", " optimal_quantities = np.array(res)\n", " return optimal_quantities" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "185.07666666666668" ] }, "execution_count": 16, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# we don't know the profit margin of each item, so we assume it is 50%.\n", "profit_margin = 0.5\n", "\n", "prices = X_test[\"sell_price\"].values\n", "stocking_cost = prices / (1 + profit_margin)\n", "\n", "# compute optimal quantities\n", "optimal_quantities = newsvendor_optimal_quantity(y_test_samples, prices, stocking_cost)\n", "\n", "# Treeffuser models continuous responses, hence we cast the predicted quantities into int\n", "optimal_quantities = optimal_quantities.astype(int)\n", "\n", "profit = newsvendor_utility(y_test, optimal_quantities, prices, stocking_cost)\n", "profit.sum()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We visualize the cumulative profit, the average demand and inventory over time in the plot below." ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " date profit avg_inventory_weighted avg_demand_weighted\n", "0 2011-11-25 -2.213333 0.039476 0.784104\n", "1 2011-11-26 1.600000 0.033068 0.933309\n", "2 2011-11-27 0.053333 0.046308 0.796895\n", "3 2011-11-28 5.873333 0.067636 0.789166\n", "4 2011-11-29 1.580000 0.041286 0.913090\n", ".. ... ... ... ...\n", "60 2012-01-24 6.336667 0.138240 0.661504\n", "61 2012-01-25 7.876667 0.078847 0.530883\n", "62 2012-01-26 -3.493333 0.051774 0.894360\n", "63 2012-01-27 -0.710000 0.102579 1.093555\n", "64 2012-01-28 11.890000 0.116262 1.258714\n", "\n", "[65 rows x 4 columns]\n" ] } ], "source": [ "performance_data = pd.DataFrame(\n", " {\n", " \"date\": dates_test,\n", " \"profit\": profit,\n", " \"demand\": y_test,\n", " \"inventory\": optimal_quantities,\n", " \"price\": prices,\n", " }\n", ")\n", "\n", "# for each day, compute average demand and inventory weighted by price\n", "daily_summary = (\n", " performance_data.groupby(\"date\")\n", " .agg(\n", " profit=(\"profit\", \"sum\"),\n", " avg_inventory_weighted=(\n", " \"inventory\",\n", " lambda x: (x * performance_data.loc[x.index, \"price\"]).sum()\n", " / performance_data.loc[x.index, \"price\"].sum(),\n", " ),\n", " avg_demand_weighted=(\n", " \"demand\",\n", " lambda x: (x * performance_data.loc[x.index, \"price\"]).sum()\n", " / performance_data.loc[x.index, \"price\"].sum(),\n", " ),\n", " )\n", " .reset_index()\n", ")\n", "\n", "print(daily_summary)" ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# dictionary to store color, alpha, linewidth, and linestyle for each line\n", "line_styles = {\n", " \"cumulative_profit\": {\"color\": \"teal\", \"alpha\": 1, \"linewidth\": 2.5, \"linestyle\": \"-\"},\n", " \"inventory\": {\n", " \"color\": \"blue\",\n", " \"alpha\": 0.5,\n", " \"linewidth\": 1.5,\n", " \"linestyle\": \"--\",\n", " },\n", " \"demand\": {\n", " \"color\": \"orange\",\n", " \"alpha\": 0.5,\n", " \"linewidth\": 1.5,\n", " \"linestyle\": \"--\",\n", " },\n", "}\n", "\n", "# create figure\n", "fig, ax1 = plt.subplots(figsize=(10, 6))\n", "\n", "# define x-axis\n", "dates = pd.to_datetime(daily_summary[\"date\"])\n", "\n", "# plot cumulative profit\n", "ax1.plot(\n", " dates,\n", " daily_summary[\"profit\"].cumsum(),\n", " **line_styles[\"cumulative_profit\"],\n", " label=\"Cumulative Profit\",\n", ")\n", "ax1.set_xlabel(\"Date\", fontsize=12)\n", "ax1.set_ylabel(\"Cumulative Profit ($)\", fontsize=12)\n", "ax1.grid(True, linestyle=\"--\", alpha=0.6)\n", "\n", "# create second y-axis for price-weighted inventory and demand\n", "ax2 = ax1.twinx()\n", "ax2.plot(\n", " dates,\n", " daily_summary[\"avg_inventory_weighted\"],\n", " **line_styles[\"inventory\"],\n", " label=\"Avg Inventory (Price Weighted)\",\n", ")\n", "ax2.plot(\n", " dates,\n", " daily_summary[\"avg_demand_weighted\"],\n", " **line_styles[\"demand\"],\n", " label=\"Avg Demand (Price Weighted)\",\n", ")\n", "ax2.set_ylabel(\"Avg Inventory and Demand (Units)\", fontsize=12)\n", "\n", "# combine all legends into one\n", "lines1, labels1 = ax1.get_legend_handles_labels()\n", "lines2, labels2 = ax2.get_legend_handles_labels()\n", "ax1.legend(\n", " lines1 + lines2, labels1 + labels2, loc=\"upper center\", bbox_to_anchor=(0.5, 1.12), ncol=3\n", ")\n", "\n", "fig.autofmt_xdate() # rotate x-tick labels\n", "\n", "plt.show()" ] } ], "metadata": { "kernelspec": { "display_name": ".venv", "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": "3.9.6" } }, "nbformat": 4, "nbformat_minor": 4 }