Skip to content

Instantly share code, notes, and snippets.

@calvinmccarter

calvinmccarter/m5-ppc-notebook-replicate-treeffuser-notuning.ipynb

Created September 27, 2024 22:13
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save calvinmccarter/2cce9f6d06adbd48efebf32d788c84b9 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save calvinmccarter/2cce9f6d06adbd48efebf32d788c84b9 to your computer and use it in GitHub Desktop.
Download ZIP
reproducing Treeffuser m5 no tuning
Display the source blob
Display the rendered blob
Raw
m5-ppc-notebook-replicate-treeffuser-notuning.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 37,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"['/Users/cmccarter/miniconda3/envs/maskingtrees/lib/python39.zip', '/Users/cmccarter/miniconda3/envs/maskingtrees/lib/python3.9', '/Users/cmccarter/miniconda3/envs/maskingtrees/lib/python3.9/lib-dynload', '', '/Users/cmccarter/.local/lib/python3.9/site-packages', '/Users/cmccarter/miniconda3/envs/maskingtrees/lib/python3.9/site-packages', '/Users/cmccarter/sandbox/treeffuser/testbed/src', '/Users/cmccarter/sandbox/treeffuser/src', '../src', '../src']\n"
]
}
],
"source": [
"import sys\n",
"sys.path.append(\"../src\")\n",
"print(sys.path)\n",
"# autoreload\n",
"import lightgbm as lgb\n",
"\n",
"#from testbed.models.quantile_regression import QuantileRegression\n",
"from testbed.models.treeffuser import Treeffuser\n",
"import pandas as pd\n",
"import numpy as np\n",
"\n",
"from pathlib import Path\n",
"from tqdm import tqdm\n",
"import pickle as pkl\n",
"\n",
"\n",
"\n",
"#from testbed.models.ngboost import NGBoostGaussian, NGBoostMixtureGaussian, NGBoostPoisson\n",
"from testbed.models.base_model import BayesOptProbabilisticModel\n",
"\n",
"\n",
"from functools import partial\n",
"\n",
"from jaxtyping import Float, Array\n",
"from typing import List, Callable\n",
"\n",
"import seaborn as sns\n",
"import matplotlib.pyplot as plt\n",
"from testbed.metrics.log_likelihood import LogLikelihoodFromSamplesMetric\n",
"from testbed.metrics.crps import CRPS\n",
"from testbed.metrics.accuracy import AccuracyMetric\n",
"\n",
"\n",
"path = \"/Users/cmccarter/sandbox/treeffuser/testbed/src/testbed/data/m5\"\n",
"\n",
"# load autoreload extension\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"# These are config variables\n",
"\n",
"PROCESS_FROM_SCRATCH = True\n",
"USE_SUBSET = True\n",
"CONTEXT_LENGTH = 20\n",
"RUN_DEPRECATED = False"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"columns of sell_prices_df:\n",
"store_id\n",
"item_id\n",
"wm_yr_wk\n",
"sell_price\n",
"\n",
"columns of sales_train_validation_df:\n",
"id\n",
"item_id\n",
"dept_id\n",
"cat_id\n",
"store_id\n",
"state_id\n",
"\n",
"columns of calendar_df:\n",
"date\n",
"wm_yr_wk\n",
"weekday\n",
"wday\n",
"month\n",
"year\n",
"d\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",
"number of zeros in sales_train_validation_df: 39777094\n"
]
}
],
"source": [
"# READ IN DATA\n",
"\n",
"sell_prices_df = pd.read_csv(Path(path) / \"sell_prices.csv\")\n",
"sales_train_validation_df = pd.read_csv(Path(path) / \"sales_train_validation.csv\")\n",
"calendar_df = pd.read_csv(Path(path) / \"calendar.csv\")\n",
"\n",
"print(\"\\ncolumns of sell_prices_df:\")\n",
"[print(col) for col in sell_prices_df.columns]\n",
"print(\"\\ncolumns of sales_train_validation_df:\")\n",
"[print(col) for col in sales_train_validation_df.columns if not col.startswith(\"d_\")]\n",
"print(\"\\ncolumns of calendar_df:\") # ommit d_1, d_2, ..., d_1913\n",
"[print(col) for col in calendar_df.columns if not col.startswith(\"d_\")]\n",
"\n",
"\"\"\n",
"\n",
"# print number of zeros\n",
"print(\"number of zeros in sales_train_validation_df: \", (sales_train_validation_df == 0).sum().sum())"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"number of zeros in sales_train_validation_df: 39777094 out of 58327370 entries\n",
"percentage of zeros in sales_train_validation_df: 68.20%\n"
]
}
],
"source": [
"#num_zeros = sales_train_validation_df.isin([0]).sum().sum()\n",
"#total_entries = sales_train_validation_df.\n",
"\n",
"items_sold_cols = sales_train_validation_df.columns[sales_train_validation_df.columns.str.startswith(\"d_\")]\n",
"num_zeros = (sales_train_validation_df[items_sold_cols] == 0).sum().sum()\n",
"total_entries = sales_train_validation_df[items_sold_cols].shape[0] * sales_train_validation_df[items_sold_cols].shape[1]\n",
"\n",
"print(f\"number of zeros in sales_train_validation_df: {num_zeros} out of {total_entries} entries\")\n",
"print(f\"percentage of zeros in sales_train_validation_df: {num_zeros / total_entries * 100:.2f}%\")"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"# 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\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Brief snapshots of the dataset"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>date</th>\n",
" <th>wm_yr_wk</th>\n",
" <th>weekday</th>\n",
" <th>wday</th>\n",
" <th>month</th>\n",
" <th>year</th>\n",
" <th>d</th>\n",
" <th>event_name_1</th>\n",
" <th>event_type_1</th>\n",
" <th>event_name_2</th>\n",
" <th>event_type_2</th>\n",
" <th>snap_CA</th>\n",
" <th>snap_TX</th>\n",
" <th>snap_WI</th>\n",
" <th>day</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>2011-01-29</td>\n",
" <td>11101</td>\n",
" <td>Saturday</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>2011</td>\n",
" <td>d_1</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>29</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>2011-01-30</td>\n",
" <td>11101</td>\n",
" <td>Sunday</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>2011</td>\n",
" <td>d_2</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>30</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>2011-01-31</td>\n",
" <td>11101</td>\n",
" <td>Monday</td>\n",
" <td>3</td>\n",
" <td>1</td>\n",
" <td>2011</td>\n",
" <td>d_3</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>31</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>2011-02-01</td>\n",
" <td>11101</td>\n",
" <td>Tuesday</td>\n",
" <td>4</td>\n",
" <td>2</td>\n",
" <td>2011</td>\n",
" <td>d_4</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>2011-02-02</td>\n",
" <td>11101</td>\n",
" <td>Wednesday</td>\n",
" <td>5</td>\n",
" <td>2</td>\n",
" <td>2011</td>\n",
" <td>d_5</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"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": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"calendar_df.head()"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>id</th>\n",
" <th>item_id</th>\n",
" <th>dept_id</th>\n",
" <th>cat_id</th>\n",
" <th>store_id</th>\n",
" <th>state_id</th>\n",
" <th>d_1</th>\n",
" <th>d_2</th>\n",
" <th>d_3</th>\n",
" <th>d_4</th>\n",
" <th>...</th>\n",
" <th>d_1904</th>\n",
" <th>d_1905</th>\n",
" <th>d_1906</th>\n",
" <th>d_1907</th>\n",
" <th>d_1908</th>\n",
" <th>d_1909</th>\n",
" <th>d_1910</th>\n",
" <th>d_1911</th>\n",
" <th>d_1912</th>\n",
" <th>d_1913</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>HOBBIES_1_001_CA_1_validation</td>\n",
" <td>HOBBIES_1_001</td>\n",
" <td>HOBBIES_1</td>\n",
" <td>HOBBIES</td>\n",
" <td>CA_1</td>\n",
" <td>CA</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>HOBBIES_1_002_CA_1_validation</td>\n",
" <td>HOBBIES_1_002</td>\n",
" <td>HOBBIES_1</td>\n",
" <td>HOBBIES</td>\n",
" <td>CA_1</td>\n",
" <td>CA</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>HOBBIES_1_003_CA_1_validation</td>\n",
" <td>HOBBIES_1_003</td>\n",
" <td>HOBBIES_1</td>\n",
" <td>HOBBIES</td>\n",
" <td>CA_1</td>\n",
" <td>CA</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>HOBBIES_1_004_CA_1_validation</td>\n",
" <td>HOBBIES_1_004</td>\n",
" <td>HOBBIES_1</td>\n",
" <td>HOBBIES</td>\n",
" <td>CA_1</td>\n",
" <td>CA</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>5</td>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>7</td>\n",
" <td>2</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>HOBBIES_1_005_CA_1_validation</td>\n",
" <td>HOBBIES_1_005</td>\n",
" <td>HOBBIES_1</td>\n",
" <td>HOBBIES</td>\n",
" <td>CA_1</td>\n",
" <td>CA</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>4</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>5 rows × 1919 columns</p>\n",
"</div>"
],
"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": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"sales_train_validation_df.head()"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>store_id</th>\n",
" <th>item_id</th>\n",
" <th>wm_yr_wk</th>\n",
" <th>sell_price</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>CA_1</td>\n",
" <td>HOBBIES_1_001</td>\n",
" <td>11325</td>\n",
" <td>9.58</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>CA_1</td>\n",
" <td>HOBBIES_1_001</td>\n",
" <td>11326</td>\n",
" <td>9.58</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>CA_1</td>\n",
" <td>HOBBIES_1_001</td>\n",
" <td>11327</td>\n",
" <td>8.26</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>CA_1</td>\n",
" <td>HOBBIES_1_001</td>\n",
" <td>11328</td>\n",
" <td>8.26</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>CA_1</td>\n",
" <td>HOBBIES_1_001</td>\n",
" <td>11329</td>\n",
" <td>8.26</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"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": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"sell_prices_df.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Process the data"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"TOTAL_ITEMS = 5000\n",
"# select a random subset of items\n",
"if USE_SUBSET:\n",
" np.random.seed(0)\n",
" unique_ids = sales_train_validation_df[\"id\"].unique()\n",
" ids = np.random.choice(sales_train_validation_df[\"id\"].unique(), TOTAL_ITEMS, replace=False)\n",
" sales_train_validation_df_sub = sales_train_validation_df[sales_train_validation_df[\"id\"].isin(ids)]\n",
" item_ids = sales_train_validation_df_sub[\"item_id\"].unique()\n",
" sell_prices_df_sub = sell_prices_df[sell_prices_df[\"item_id\"].isin(item_ids)]\n",
" calendar_df_sub = calendar_df\n",
"\n",
"\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"columns_sales_train_validation.head)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The strategy for processing the data is going to be the following. 1) We are going to have X and y where y is the next days sales for a given product. 3) X is made up of 10 previous prices, day of the week, + event types, cat_id, store_id, state_id"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"def proc_train_test(sales_train_validation_df: pd.DataFrame, calendar_df: pd.DataFrame, sell_prices_df: pd.DataFrame, context_length: int, test_percentage: float, percentage_omittied: int = 0): #type annotation too long\n",
" \"\"\"\n",
" This function processes the data and returns the training and test data in two ways:\n",
" - undifferentiated: a list of all training and test data (X_train, y_train, X_test, y_test)\n",
" - differentiated: a list of training and test data for each product (X_train_prod, y_train_prod, X_test_prod, y_test_prod)\n",
" where X_train_prod[i] contains a list of all X_train values for the product i with similar grouping for y_train_prod and test\n",
"\n",
" This assumes from the dataframes that\n",
" - sales_train_validation_df:\n",
" - has columns with the format d_1, d_2, ...\n",
" - has columns item_id and store_id\n",
" - calendar_df:\n",
" - wday, month, event_name_1, event_name_2\n",
" - sell_prices_df:\n",
" - item_id, store_id, sell_price\n",
"\n",
" - percentage_omittied: percentage of the data to be omitted from the training data and the test data\n",
" (randomly selected)\n",
"\n",
" Returns:\n",
" - undifferentiated: Tuple of X_train, y_train, X_test, y_test\n",
" - differentiated: Tuple of X_train_prod, y_train_prod, X_test_prod, y_test_prod\n",
" \"\"\"\n",
" np.random.seed(0)\n",
" # First we need to get the training data\n",
" # We will use the first 1913 days as training data and the next\n",
"\n",
" X_train = []\n",
" y_train = []\n",
"\n",
" X_test = []\n",
" y_test = []\n",
"\n",
" # We will also return a second grouping of lists where X_train_prod[i] contains a\n",
" # a list of all X_train values for the product i with similar grouping for y_train_prod and test\n",
" X_train_prod = []\n",
" y_train_prod = []\n",
" X_test_prod = []\n",
" y_test_prod = []\n",
"\n",
"\n",
" # get all days that start with d_ and look for the maximum\n",
" total_days = max([int(x.split(\"_\")[1]) for x in sales_train_validation_df.columns if \"d_\" in x])\n",
" train_days = int(total_days * (1 - test_percentage))\n",
" print(\"train days\", train_days)\n",
" print(\"test days\", total_days - train_days)\n",
" print(\"total days\", total_days)\n",
"\n",
" # Precompute the required data\n",
" calendar_df_dict = calendar_df.set_index(\"d\").to_dict(orient=\"index\")\n",
" sell_prices_dict = sell_prices_df.groupby([\"item_id\", \"store_id\"])[\"sell_price\"].first().to_dict()\n",
"\n",
" pbar = tqdm(total=len(sales_train_validation_df))\n",
" for _, row in sales_train_validation_df.iterrows():\n",
" item_id = row[\"item_id\"]\n",
" store_id = row[\"store_id\"]\n",
"\n",
" X_train_prod.append([])\n",
" y_train_prod.append([])\n",
" X_test_prod.append([])\n",
" y_test_prod.append([])\n",
"\n",
" pbar.update(1)\n",
"\n",
" valid_size = int((train_days - context_length) * (1 - percentage_omittied))\n",
" valid_js = np.random.choice(range(1, train_days - context_length), valid_size, replace=False)\n",
"\n",
" valid_js = list(valid_js) + list(range(train_days, total_days - context_length))\n",
"\n",
" for j in valid_js:\n",
" x = []\n",
"\n",
" # Add sales values for the previous context_length days\n",
" x.extend(row[f\"d_{j+k}\"] for k in range(context_length))\n",
"\n",
" # Add additional features\n",
" current_day = f\"d_{j+context_length}\"\n",
" calendar_data = calendar_df_dict[current_day]\n",
" x.extend([\n",
" calendar_data[\"wday\"],\n",
" calendar_data[\"month\"],\n",
" store_id,\n",
" calendar_data[\"event_name_1\"],\n",
" calendar_data[\"event_name_2\"],\n",
" sell_prices_dict[(item_id, store_id)],\n",
" item_id\n",
" ])\n",
"\n",
" if j < train_days:\n",
" X_train.append(x)\n",
" y_train.append(row[current_day])\n",
" X_train_prod[-1].append(x)\n",
" y_train_prod[-1].append(row[current_day])\n",
"\n",
" else:\n",
" X_test.append(x)\n",
" y_test.append(row[current_day])\n",
" X_train_prod[-1].append(x)\n",
" y_train_prod[-1].append(row[current_day])\n",
"\n",
" undifferentiated = (X_train, y_train, X_test, y_test)\n",
" differentiated = (X_train_prod, y_train_prod, X_test_prod, y_test_prod)\n",
" return undifferentiated, differentiated"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"train days 1874\n",
"test days 39\n",
"total days 1913\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|█████████████████████████████████████| 5000/5000 [00:04<00:00, 1213.90it/s]\n"
]
}
],
"source": [
"if PROCESS_FROM_SCRATCH:\n",
" undifferentiated, differentiated = proc_train_test(sales_train_validation_df_sub, calendar_df, sell_prices_df_sub, CONTEXT_LENGTH, 0.02, 0.99)\n",
" X_train, y_train, X_test, y_test = undifferentiated\n",
" X_train_prod, y_train_prod, X_test_prod, y_test_prod = differentiated\n"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(90000, 90000, 95000, 95000)"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"len(X_train), len(y_train), len(X_test), len(y_test)"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [],
"source": [
"COL_NAMES = [\n",
" f\"day_{i}\" for i in range(1, CONTEXT_LENGTH+1)\n",
"] + [\"wday\", \"month\", \"store_id\", \"event_name_1\", \"event_name_2\", \"sell_price\", \"item_id\"]\n",
"\n",
"CAT_COLS = [\"store_id\", \"event_name_1\", \"event_name_2\", \"item_id\", \"wday\", \"month\"]\n",
"CAT_COLS_IDX = [COL_NAMES.index(col) for col in CAT_COLS]\n"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
"X_train_df = pd.DataFrame(X_train)\n",
"X_test_df = pd.DataFrame(X_test)\n",
"y_test_df = pd.DataFrame(y_test)\n",
"y_train_df = pd.DataFrame(y_train)\n",
"\n",
"X_train_df.columns = COL_NAMES\n",
"X_test_df.columns = COL_NAMES"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>day_1</th>\n",
" <th>day_2</th>\n",
" <th>day_3</th>\n",
" <th>day_4</th>\n",
" <th>day_5</th>\n",
" <th>day_6</th>\n",
" <th>day_7</th>\n",
" <th>day_8</th>\n",
" <th>day_9</th>\n",
" <th>day_10</th>\n",
" <th>...</th>\n",
" <th>day_18</th>\n",
" <th>day_19</th>\n",
" <th>day_20</th>\n",
" <th>wday</th>\n",
" <th>month</th>\n",
" <th>store_id</th>\n",
" <th>event_name_1</th>\n",
" <th>event_name_2</th>\n",
" <th>sell_price</th>\n",
" <th>item_id</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>6</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>CA_1</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>2.98</td>\n",
" <td>HOBBIES_1_005</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>3</td>\n",
" <td>3</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>3</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>3</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>4</td>\n",
" <td>10</td>\n",
" <td>CA_1</td>\n",
" <td>EidAlAdha</td>\n",
" <td>NaN</td>\n",
" <td>2.98</td>\n",
" <td>HOBBIES_1_005</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>...</td>\n",
" <td>0</td>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>CA_1</td>\n",
" <td>SuperBowl</td>\n",
" <td>NaN</td>\n",
" <td>2.98</td>\n",
" <td>HOBBIES_1_005</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>...</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>4</td>\n",
" <td>4</td>\n",
" <td>CA_1</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>2.98</td>\n",
" <td>HOBBIES_1_005</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>4</td>\n",
" <td>3</td>\n",
" <td>CA_1</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>2.98</td>\n",
" <td>HOBBIES_1_005</td>\n",
" </tr>\n",
" <tr>\n",
" <th>...</th>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>89995</th>\n",
" <td>5</td>\n",
" <td>5</td>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>4</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>1</td>\n",
" <td>9</td>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>5</td>\n",
" <td>WI_3</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>1.00</td>\n",
" <td>FOODS_3_827</td>\n",
" </tr>\n",
" <tr>\n",
" <th>89996</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>WI_3</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>1.00</td>\n",
" <td>FOODS_3_827</td>\n",
" </tr>\n",
" <tr>\n",
" <th>89997</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>WI_3</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>1.00</td>\n",
" <td>FOODS_3_827</td>\n",
" </tr>\n",
" <tr>\n",
" <th>89998</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>8</td>\n",
" <td>WI_3</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>1.00</td>\n",
" <td>FOODS_3_827</td>\n",
" </tr>\n",
" <tr>\n",
" <th>89999</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>11</td>\n",
" <td>WI_3</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>1.00</td>\n",
" <td>FOODS_3_827</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>90000 rows × 27 columns</p>\n",
"</div>"
],
"text/plain": [
" day_1 day_2 day_3 day_4 day_5 day_6 day_7 day_8 day_9 day_10 \\\n",
"0 2 0 0 4 1 0 6 0 0 0 \n",
"1 3 3 1 1 1 0 0 3 0 0 \n",
"2 0 0 1 2 1 1 1 1 1 1 \n",
"3 0 2 0 1 3 2 1 1 2 2 \n",
"4 0 0 0 0 0 0 0 0 0 0 \n",
"... ... ... ... ... ... ... ... ... ... ... \n",
"89995 5 5 0 2 3 0 1 4 2 0 \n",
"89996 0 0 0 0 0 0 0 0 0 0 \n",
"89997 0 0 0 0 0 0 0 0 0 0 \n",
"89998 0 0 0 0 0 0 0 0 0 0 \n",
"89999 0 0 0 0 0 0 0 0 0 0 \n",
"\n",
" ... day_18 day_19 day_20 wday month store_id event_name_1 \\\n",
"0 ... 1 2 0 2 2 CA_1 NaN \n",
"1 ... 3 0 1 4 10 CA_1 EidAlAdha \n",
"2 ... 0 4 1 2 2 CA_1 SuperBowl \n",
"3 ... 2 0 2 4 4 CA_1 NaN \n",
"4 ... 0 0 0 4 3 CA_1 NaN \n",
"... ... ... ... ... ... ... ... ... \n",
"89995 ... 1 9 0 2 5 WI_3 NaN \n",
"89996 ... 0 0 0 1 3 WI_3 NaN \n",
"89997 ... 0 0 0 4 1 WI_3 NaN \n",
"89998 ... 0 0 0 2 8 WI_3 NaN \n",
"89999 ... 0 0 0 2 11 WI_3 NaN \n",
"\n",
" event_name_2 sell_price item_id \n",
"0 NaN 2.98 HOBBIES_1_005 \n",
"1 NaN 2.98 HOBBIES_1_005 \n",
"2 NaN 2.98 HOBBIES_1_005 \n",
"3 NaN 2.98 HOBBIES_1_005 \n",
"4 NaN 2.98 HOBBIES_1_005 \n",
"... ... ... ... \n",
"89995 NaN 1.00 FOODS_3_827 \n",
"89996 NaN 1.00 FOODS_3_827 \n",
"89997 NaN 1.00 FOODS_3_827 \n",
"89998 NaN 1.00 FOODS_3_827 \n",
"89999 NaN 1.00 FOODS_3_827 \n",
"\n",
"[90000 rows x 27 columns]"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"\n",
"X_train_df"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>day_1</th>\n",
" <th>day_2</th>\n",
" <th>day_3</th>\n",
" <th>day_4</th>\n",
" <th>day_5</th>\n",
" <th>day_6</th>\n",
" <th>day_7</th>\n",
" <th>day_8</th>\n",
" <th>day_9</th>\n",
" <th>day_10</th>\n",
" <th>...</th>\n",
" <th>day_18</th>\n",
" <th>day_19</th>\n",
" <th>day_20</th>\n",
" <th>wday</th>\n",
" <th>month</th>\n",
" <th>store_id</th>\n",
" <th>event_name_1</th>\n",
" <th>event_name_2</th>\n",
" <th>sell_price</th>\n",
" <th>item_id</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>6</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>30</td>\n",
" <td>4</td>\n",
" <td>2.98</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>3</td>\n",
" <td>3</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>3</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>3</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>9</td>\n",
" <td>0</td>\n",
" <td>6</td>\n",
" <td>4</td>\n",
" <td>2.98</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>...</td>\n",
" <td>0</td>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>26</td>\n",
" <td>4</td>\n",
" <td>2.98</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>...</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>3</td>\n",
" <td>0</td>\n",
" <td>30</td>\n",
" <td>4</td>\n",
" <td>2.98</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>...</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>30</td>\n",
" <td>4</td>\n",
" <td>2.98</td>\n",
" <td>0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>5 rows × 27 columns</p>\n",
"</div>"
],
"text/plain": [
" day_1 day_2 day_3 day_4 day_5 day_6 day_7 day_8 day_9 day_10 ... \\\n",
"0 2 0 0 4 1 0 6 0 0 0 ... \n",
"1 3 3 1 1 1 0 0 3 0 0 ... \n",
"2 0 0 1 2 1 1 1 1 1 1 ... \n",
"3 0 2 0 1 3 2 1 1 2 2 ... \n",
"4 0 0 0 0 0 0 0 0 0 0 ... \n",
"\n",
" day_18 day_19 day_20 wday month store_id event_name_1 event_name_2 \\\n",
"0 1 2 0 1 1 0 30 4 \n",
"1 3 0 1 3 9 0 6 4 \n",
"2 0 4 1 1 1 0 26 4 \n",
"3 2 0 2 3 3 0 30 4 \n",
"4 0 0 0 3 2 0 30 4 \n",
"\n",
" sell_price item_id \n",
"0 2.98 0 \n",
"1 2.98 0 \n",
"2 2.98 0 \n",
"3 2.98 0 \n",
"4 2.98 0 \n",
"\n",
"[5 rows x 27 columns]"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Encode the categorical columns as numbers\n",
"from sklearn.preprocessing import LabelEncoder\n",
"# Get only label of item_id\n",
"X_train_df[\"item_id\"] = X_train_df[\"item_id\"].apply(lambda x: x.split(\"_\")[1])\n",
"X_test_df[\"item_id\"] = X_test_df[\"item_id\"].apply(lambda x: x.split(\"_\")[1])\n",
"\n",
"\n",
"label_encoders = {}\n",
"for col in CAT_COLS:\n",
" le = LabelEncoder()\n",
" X_train_df[col] = le.fit_transform(X_train_df[col])\n",
" X_test_df[col] = le.transform(X_test_df[col])\n",
" label_encoders[col] = le\n",
"\n",
"\n",
"X_train_prod_processed = []\n",
"X_test_prod_processed = []\n",
"for i in range(len(X_train_prod)):\n",
" X_train_prod_processed.append(pd.DataFrame(X_train_prod[i], columns=COL_NAMES))\n",
" X_test_prod_processed.append(pd.DataFrame(X_test_prod[i], columns=COL_NAMES))\n",
" X_train_prod_processed[-1][\"item_id\"] = X_train_prod_processed[-1][\"item_id\"].apply(lambda x: x.split(\"_\")[1])\n",
" X_test_prod_processed[-1][\"item_id\"] = X_test_prod_processed[-1][\"item_id\"].apply(lambda x: x.split(\"_\")[1])\n",
" for col in CAT_COLS:\n",
" X_train_prod_processed[-1][col] = label_encoders[col].transform(X_train_prod_processed[-1][col])\n",
" X_test_prod_processed[-1][col] = label_encoders[col].transform(X_test_prod_processed[-1][col])\n",
"\n",
"X_train_df.head()\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# PPC"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### \"Standard PPCs\""
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [],
"source": [
"\n",
"def max_ppc(y_true: Float[Array, \"batch y_dim\"], y_samples: Float[Array, \"samples batch y_dim\"], number=0, name=\"\") -> None:\n",
" # rpeat y_true to match the shape of y_samples\n",
" max_ppc = np.max(y_samples, axis=1)\n",
" true_max = np.max(y_true)\n",
"\n",
" return max_ppc.flatten(), true_max.flatten(), \"max_ppc\"\n",
"\n",
"def quantile_ppc(y_true: Float[Array, \"batch y_dim\"], y_samples: Float[Array, \"samples batch y_dim\"], quantile=0.5, number=0, name=\"\") -> None:\n",
" # rpeat y_true to match the shape of y_samples\n",
" q = np.quantile(y_samples, quantile, axis=1)\n",
" true_q = np.quantile(y_true, quantile)\n",
" return q.flatten(), true_q.flatten(), f\"quantile_ppc_{quantile}\"\n",
"\n",
"def zeros(y_true: Float[Array, \"batch y_dim\"], y_samples: Float[Array, \"samples batch y_dim\"], number=0, name=\"\") -> None:\n",
" \"Count the number of zeros in the samples\"\n",
" zeros = np.sum(y_samples < 0.1, axis=1)\n",
" true_zeros = np.sum(y_true < 0.1)\n",
"\n",
" return zeros.flatten(), true_zeros.flatten(), \"zeros\"\n",
"\n",
"def percentage_zeros(y_true: Float[Array, \"batch y_dim\"], y_samples: Float[Array, \"samples batch y_dim\"], number=0, name=\"\") -> None:\n",
" \"Count the number of zeros in the samples\"\n",
" zeros = np.mean(y_samples < 0.1, axis=1)\n",
" true_zeros = np.mean(y_true < 0.1)\n",
"\n",
" return zeros.flatten(), true_zeros.flatten(), \"percentage_zeros\""
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [],
"source": [
"def plot_ppcs(y_true: Float[Array, \"batch y_dim\"], y_samples: Float[Array, \"samples batch y_dim\"], ppcs: List[Callable],\n",
" number=0, name=\"\") -> None:\n",
" # plot the distribution of\n",
"\n",
" for ppc in ppcs:\n",
" ppc(y_true, y_samples, number=number, name=name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### \"Complex PPCs\""
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [],
"source": [
"\n",
"def plot_model_comparisons(data, y_true, figsize=(12, 8), model_names=None):\n",
" \"\"\"\n",
" Plots model predictions against true values for each day.\n",
"\n",
" :param data: numpy array of shape [models, samples, days] containing model predictions\n",
" :param y_true: array of shape [days] containing the true values\n",
" :param figsize: tuple indicating the size of the figure\n",
" \"\"\"\n",
" sns.set(style=\"whitegrid\")\n",
" models, samples, days = data.shape\n",
"\n",
" # Create a figure and axis object\n",
" fig, ax = plt.subplots(figsize=figsize)\n",
"\n",
" # We will transform the data to a format suitable for seaborn\n",
" # Create a DataFrame with model, day, and sample values\n",
" plot_data = []\n",
" if model_names is None:\n",
" model_names = [f\"Model {i}\" for i in range(models)]\n",
"\n",
" for model_idx in range(models):\n",
" for day_idx in range(days):\n",
" for sample_idx in range(samples):\n",
" plot_data.append({\n",
" \"Day\": day_idx,\n",
" \"Value\": data[model_idx, sample_idx, day_idx],\n",
" \"Model\": model_names[model_idx]\n",
" })\n",
"\n",
" import pandas as pd\n",
" plot_data = pd.DataFrame(plot_data)\n",
"\n",
" # Use seaborn to plot the boxplots\n",
" sns.boxplot(x=\"Day\", y=\"Value\", hue=\"Model\", data=plot_data, ax=ax, width=0.6)\n",
"\n",
" # Plot true values\n",
" plt.plot(y_true, 'o', color='red', label='True Values')\n",
"\n",
" # Setting labels and title\n",
" plt.xticks(ticks=np.arange(days), labels=[f\"Day {i+1}\" for i in range(days)])\n",
" plt.xlabel('Days')\n",
" plt.ylabel('Values')\n",
" plt.title('Model Predictions vs. True Values')\n",
" plt.legend()\n",
"\n",
" # Show the plot\n",
" plt.show()\n",
"\n",
"\n",
"\n",
"\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Model Evaluation"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [],
"source": [
"def save_results_to_pkl(results: dict, dir_name, name):\n",
" if not Path(dir_name).exists():\n",
" Path(dir_name).mkdir(parents=True)\n",
"\n",
" path = Path(dir_name) / f\"{name}.pkl\"\n",
" with open(path, \"wb\") as f:\n",
" pkl.dump(results, f)\n",
"\n",
"\n",
"def load_results_from_pkl(dir_name, name):\n",
" path = Path(dir_name) / f\"{name}.pkl\"\n",
" with open(path, \"rb\") as f:\n",
" results = pkl.load(f)\n",
" return results\n"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [],
"source": [
"# Simple helper function to train a model and plot ppcs\n",
"\n",
"def get_ppcs(y_samples, X_test, y_test, ppcs, number=0, name=\"\") -> None:\n",
" \"\"\"\n",
" Returns a dictionary with the samples and the true values for each ppc\n",
" the dictionary a\n",
" \"\"\"\n",
" y_samples = np.array(y_samples)\n",
" #y_samples = np.maximum(y_samples, 0)\n",
" #y_samples = np.round(y_samples, 0)\n",
"\n",
" ppc_results = {}\n",
" for ppc in ppcs:\n",
" samples, true, name = ppc(y_test, y_samples, number=number, name=name)\n",
" ppc_results[name] = {\"samples\": samples, \"true\": true}\n",
"\n",
" return ppc_results\n"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" day_1 day_2 day_3 day_4 day_5 day_6 day_7 day_8 day_9 day_10 ... \\\n",
"0 2 0 0 4 1 0 6 0 0 0 ... \n",
"1 3 3 1 1 1 0 0 3 0 0 ... \n",
"2 0 0 1 2 1 1 1 1 1 1 ... \n",
"3 0 2 0 1 3 2 1 1 2 2 ... \n",
"4 0 0 0 0 0 0 0 0 0 0 ... \n",
"\n",
" day_18 day_19 day_20 wday month store_id event_name_1 event_name_2 \\\n",
"0 1 2 0 1 1 0 30 4 \n",
"1 3 0 1 3 9 0 6 4 \n",
"2 0 4 1 1 1 0 26 4 \n",
"3 2 0 2 3 3 0 30 4 \n",
"4 0 0 0 3 2 0 30 4 \n",
"\n",
" sell_price item_id \n",
"0 2.98 0 \n",
"1 2.98 0 \n",
"2 2.98 0 \n",
"3 2.98 0 \n",
"4 2.98 0 \n",
"\n",
"[5 rows x 27 columns]\n"
]
}
],
"source": [
"print(X_train_df.head())"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [],
"source": [
"EVAL_VALUES = 10_000 #len(X_test_df)\n",
"np.random.seed(0)\n",
"\n",
"eval_idx = np.random.choice(len(X_test_df), EVAL_VALUES, replace=False)\n",
"\n",
"X_train_np = X_train_df.values\n",
"X_test_np = X_test_df.values[eval_idx]\n",
"\n",
"y_train_np = y_train_df.values + np.random.normal(0, 0.01, y_train_df.shape)\n",
"y_test_np = y_test_df.values[eval_idx]\n",
"\n",
"# change to float to prevent errors\n",
"y_train_np = y_train_np.astype(np.float32)\n",
"y_test_np = y_test_np.astype(np.float32)\n",
"\n",
"dataset = {\n",
" \"X_train\": X_train_np,\n",
" \"X_test\": X_test_np,\n",
" \"y_train\": y_train_np,\n",
" \"y_test\": y_test_np,\n",
" \"col_names\": COL_NAMES,\n",
" \"cat_cols\": CAT_COLS,\n",
"}\n",
"\n",
"with open(\"dataset.pkl\", \"wb\") as f:\n",
" pkl.dump(dataset, f)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {},
"outputs": [],
"source": [
"\n",
"#MODEL_CLASSES = MODEL_CLASSES[-1:]\n",
"#NAMES = NAMES[-1:]\n",
"\n",
"NUM_SAMPLES = 100\n",
"HYPERS = [\n",
" {\"subsample\": 0.20, \"subsample_freq\": 1, \"verbose\": 0, \"num_leaves\":129, \"learning_rate\":0.5,\n",
" \"sde_manual_hyperparams\": {\"hyperparam_max\": 10}},\n",
" {},\n",
" {}\n",
"]\n",
"MODEL_CLASSES = [Treeffuser]\n",
"NAMES = [\"Treeffuser\"]\n",
"HYPERS = [\n",
" {\"n_estimators\": 3000, \"learning_rate\": 0.1, \"num_leaves\": 32, \"early_stopping_rounds\": 50, \"n_repeats\": 10},\n",
"]\n",
"\n",
"results = []\n",
"for i in range(len(MODEL_CLASSES)):\n",
" model_cls = MODEL_CLASSES[i]\n",
" #model = BayesOptProbabilisticModel(model_cls, n_iter_bayes_opt=20, frac_validation=0.01)\n",
" model = model_cls(**HYPERS[i])\n",
"\n",
"\n",
" if False and model_cls == NGBoostPoisson:\n",
" # shuffle the data\n",
" np.random.seed(0)\n",
" idx = np.random.permutation(len(X_train_np))\n",
" X_train_np_ngb = X_train_np[idx]\n",
" y_train_np_ngb = y_train_np[idx].astype(np.int32)\n",
" model.fit(X_train_np_ngb, y_train_np_ngb)\n",
"\n",
" elif False and model_cls == NGBoostGaussian:\n",
" y_train_np_ngb = y_train_np + np.random.normal(0, 3, y_train_np.shape)\n",
" # rescale\n",
" #y_train_np_ngb = (y_train_np_ngb - np.mean(y_train_np_ngb)) / np.std(y_train_np_ngb)\n",
" model.fit(X_train_np, y_train_np_ngb)\n",
"\n",
" else:\n",
" model.fit(X_train_np, y_train_np)\n",
"\n",
" results.append({\n",
" \"model\": model,\n",
" \"model_name\": NAMES[i]\n",
" })\n",
"\n",
" save_results_to_pkl(results, \"m5\", \"results.pkl\")\n",
"\n",
"\n",
"results = load_results_from_pkl(\"m5\", \"results.pkl\")\n",
"\n",
"\n",
"# Save the results\n"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Model: Treeffuser\n"
]
}
],
"source": [
"\n",
"for i, result in enumerate(results):\n",
" model = result[\"model\"]\n",
" model_name = result[\"model_name\"]\n",
" print(f\"Model: {model_name}\")\n",
" y_samples = model.sample(X_test_np, NUM_SAMPLES)\n",
" results[i][\"y_samples\"] = y_samples"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {},
"outputs": [],
"source": [
"results_no_model = []\n",
"for result in results:\n",
" results_no_model.append({k: v for k, v in result.items() if k != \"model\"})\n",
"\n",
"# Don't uncomment or will overwrite the results\n",
"#save_results_to_pkl(results, \"m5\", \"results_final.pkl\")\n",
"#save_results_to_pkl(results_no_model, \"m5\", \"results_no_model_final.pkl\")"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {},
"outputs": [],
"source": [
"try:\n",
" results_no_model = load_results_from_pkl(\"m5\", \"results_final.pkl\")\n",
"except:\n",
" results_no_model = results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we can actually fit some of the models"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0\n"
]
}
],
"source": [
"ppcs = [max_ppc, zeros, percentage_zeros] + [partial(quantile_ppc, quantile=q) for q in [0.1, 0.5, 0.9, 0.99, 0.999]]\n",
"\n",
"for i, model_cls in enumerate(MODEL_CLASSES):\n",
" print(i)\n",
" ppc_results = get_ppcs(\n",
" y_samples=results_no_model[i][\"y_samples\"],\n",
" X_test=X_test_np,\n",
" y_test=y_test_np,\n",
" ppcs=ppcs,\n",
" number=i,\n",
" name=model_cls.__name__\n",
" )\n",
" results[i][\"ppc_results\"] = ppc_results\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Plot the PPCs"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {},
"outputs": [],
"source": [
"# Titles for plots\n",
"ppc_tiles = {\n",
" \"max_ppc\": \"$\\max$\",\n",
" \"zeros\": r\"$\\text{zeros}$\",\n",
" \"quantile_ppc_0.99\": \"$q_{0.99}$\",\n",
" }"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {},
"outputs": [],
"source": [
"def set_plot_style():\n",
" \"\"\"\n",
" Sets a common plotting style for all of the figures that will be\n",
" used in the final paper.\n",
" \"\"\"\n",
"\n",
" # no grid but white with pretty ticks\n",
"\n",
"\n",
" # use latex font by default\n",
" plt.rc(\"text\", usetex=False)\n",
" plt.rc(\"font\", family=\"serif\")\n",
" sns.set_style(\"white\")\n",
"\n",
" # make it ready for a presentation\n",
" sns.set_context(\"talk\")\n",
"set_plot_style()"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"samples [204. 182. 267. 127. 123. 134. 205. 223. 223. 194. 191. 184. 101. 105.\n",
" 155. 166. 194. 130. 122. 139. 147. 216. 220. 144. 183. 237. 147. 182.\n",
" 156. 108. 243. 247. 118. 154. 184. 114. 111. 141. 190. 289. 140. 254.\n",
" 159. 241. 187. 119. 167. 222. 152. 96. 140. 177. 178. 137. 200. 233.\n",
" 152. 292. 239. 145. 263. 186. 112. 172. 219. 168. 136. 132. 162. 159.\n",
" 247. 109. 154. 120. 186. 166. 172. 181. 149. 140. 204. 146. 247. 204.\n",
" 269. 148. 124. 133. 101. 170. 150. 275. 138. 117. 169. 110. 345. 161.\n",
" 108. 104.]\n",
"title max_ppc\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"samples [5693 5628 5649 5628 5680 5641 5678 5637 5657 5694 5597 5599 5685 5607\n",
" 5676 5528 5707 5590 5655 5630 5618 5563 5675 5668 5624 5655 5671 5625\n",
" 5647 5610 5656 5634 5696 5759 5741 5684 5648 5617 5632 5564 5616 5656\n",
" 5635 5625 5682 5596 5696 5661 5726 5645 5644 5622 5660 5675 5653 5671\n",
" 5682 5702 5670 5640 5562 5666 5645 5703 5690 5566 5693 5641 5659 5676\n",
" 5587 5644 5641 5669 5682 5612 5632 5644 5670 5684 5643 5665 5679 5666\n",
" 5666 5657 5695 5681 5680 5661 5679 5684 5681 5686 5707 5701 5682 5606\n",
" 5652 5636]\n",
"title zeros\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"samples [1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
" 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
" 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
" 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
" 1. 1. 1. 1.]\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"samples [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
" 0. 0. 0. 0.]\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"samples [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
" 0. 0. 0. 0.]\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"samples [3. 3. 3. 3. 4. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3.\n",
" 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3.\n",
" 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3.\n",
" 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3.\n",
" 3. 3. 3. 3.]\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"samples [15. 14. 14. 14. 13. 13. 14. 14. 15. 14. 14. 15. 13. 14. 14. 15. 14. 14.\n",
" 14. 14. 14. 14. 13. 14. 13. 13. 15. 15. 14. 14. 14. 14. 14. 15. 14. 13.\n",
" 14. 14. 15. 13. 13. 13. 14. 14. 14. 14. 14. 14. 14. 15. 14. 14. 13. 15.\n",
" 14. 14. 14. 13. 15. 14. 14. 15. 14. 14. 14. 14. 14. 14. 14. 15. 14. 14.\n",
" 14. 14. 13. 14. 14. 15. 14. 14. 14. 13. 13. 14. 14. 14. 14. 15. 15. 13.\n",
" 14. 13. 13. 14. 14. 13. 14. 14. 15. 14.]\n",
"title quantile_ppc_0.99\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"samples [33. 40. 36. 32. 39. 39. 39. 35. 38. 39. 39. 39. 37. 45. 35. 42. 37. 41.\n",
" 42. 34. 38. 34. 40. 40. 45. 39. 39. 37. 37. 42. 35. 38. 36. 38. 36. 38.\n",
" 37. 37. 41. 42. 36. 37. 40. 42. 37. 38. 32. 39. 40. 36. 34. 39. 38. 40.\n",
" 40. 40. 40. 37. 36. 34. 37. 35. 42. 37. 48. 38. 35. 37. 42. 41. 42. 42.\n",
" 37. 34. 41. 34. 34. 41. 39. 37. 33. 36. 39. 40. 38. 38. 35. 38. 36. 38.\n",
" 46. 32. 34. 41. 40. 37. 41. 37. 38. 39.]\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"def proc_title(title):\n",
" if title in ppc_tiles:\n",
" print(\"title\", title)\n",
" return ppc_tiles[title]\n",
"\n",
" x = title.replace(\"_\", \" \").capitalize()\n",
" x = x.replace(\"ppc\", \"\")\n",
" return x\n",
"\n",
"ppc_number = len(ppcs)\n",
"ppc_names = results[0][\"ppc_results\"].keys()\n",
"\n",
"for ppc_name in ppc_names:\n",
" fig, ax = plt.subplots()\n",
" for i, res in enumerate(results):\n",
" model_name = res[\"model_name\"]\n",
" samples = res[\"ppc_results\"][ppc_name][\"samples\"]\n",
" # make int\n",
" samples = np.maximum(samples, 0)\n",
" samples = np.round(samples)\n",
" true = res[\"ppc_results\"][ppc_name][\"true\"]\n",
" n_unique_samples = len(np.unique(samples))\n",
" discrete = n_unique_samples < 20\n",
"\n",
" print(\"samples\", samples)\n",
"\n",
" # plot a histogram of the samples but with integers (use nice binning)\n",
" if ppc_name == \"max_ppc\":\n",
" binwidth = 70\n",
" else:\n",
" binwidth = None\n",
" sns.histplot(samples, ax=ax, label=f\"{model_name}\" + r\" $\\hat{p}$\", discrete=discrete, stat=\"density\", binwidth=binwidth)\n",
" if i == 0:\n",
" ax.axvline(true, color=\"red\", label=\"Observed Value\")\n",
" ax.set_title(proc_title(f\"{ppc_name}\"))\n",
" ax.legend()\n",
"\n",
" if n_unique_samples < 2:\n",
" min_val = np.min(samples) - 1\n",
" max_val = np.max(samples) + 1\n",
"\n",
" ax.set_xlim(min_val, max_val)\n",
"\n",
"\n",
"\n",
" # save the figure\n",
" fig.savefig(f\"m5/{ppc_name}.png\", dpi=100)\n",
" # save as pdf\n",
" fig.savefig(f\"m5/{ppc_name}.pdf\")\n",
"\n",
" #max_x = true * 5\n",
" #max_samples = np.max(samples)\n",
" #if max_samples > max_x:\n",
" # ax.set_xlim(0, max_x)\n",
"\n",
"\n",
" plt.show()\n",
"\n",
"\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 33,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.legend.Legend at 0x3acbe1d00>"
]
},
"execution_count": 33,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"for i, result in enumerate(results):\n",
" samples = result[\"y_samples\"][0]\n",
" samples = np.round(samples) + (i+1)/4\n",
" sns.histplot(samples.flatten(), stat=\"density\", label=result[\"model_name\"])\n",
"\n",
"sns.histplot(y_test_np.flatten(), stat=\"density\", color=\"red\", label=\"true\")\n",
"plt.xlim(0, 10)\n",
"plt.legend()\n"
]
},
{
"cell_type": "code",
"execution_count": 34,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"mean shape (10,)\n",
"lower shape (10,)\n",
"upper shape (10,)\n",
"y_true shape (10,)\n",
"samples shape (1, 100, 10)\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 1200x800 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"mean shape (10,)\n",
"lower shape (10,)\n",
"upper shape (10,)\n",
"y_true shape (10,)\n",
"samples shape (1, 100, 10)\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 1200x800 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"mean shape (10,)\n",
"lower shape (10,)\n",
"upper shape (10,)\n",
"y_true shape (10,)\n",
"samples shape (1, 100, 10)\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 1200x800 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"np.random.seed(0)\n",
"\n",
"NUM_PRODS_TO_PLOT = 3\n",
"DAYS_TO_PLOT = 10\n",
"QUATILE = 0.9\n",
"NUM_SAMPLES = 100\n",
"\n",
"prods_to_plot = np.random.choice(range(len(X_train_prod_processed)), NUM_PRODS_TO_PLOT)\n",
"\n",
"means_to_plot = []\n",
"lower_q_s = []\n",
"upper_q_s = []\n",
"\n",
"prod_dict = {}\n",
"\n",
"for res in results:\n",
" model = res[\"model\"]\n",
"\n",
" for i in prods_to_plot:\n",
" if i not in prod_dict:\n",
" prod_dict[i] = {\n",
" \"means_to_plot\": [],\n",
" \"lower_q_s\": [],\n",
" \"upper_q_s\": [],\n",
" \"samples\": []\n",
" }\n",
"\n",
" X_prod_proc_i = X_train_prod_processed[i][-DAYS_TO_PLOT:]\n",
" y_prod_proc_i = y_train_prod[i][-DAYS_TO_PLOT:]\n",
"\n",
" samples = model.sample(X_prod_proc_i.values, NUM_SAMPLES)\n",
" samples = samples.astype(int)\n",
" samples = np.maximum(samples, 0)\n",
"\n",
" means = np.mean(samples, axis=0)\n",
" lower_q = np.quantile(samples, 1-QUATILE, axis=0)\n",
" upper_q = np.quantile(samples, QUATILE, axis=0)\n",
"\n",
" prod_dict[i][\"means_to_plot\"].append(means)\n",
" prod_dict[i][\"lower_q_s\"].append(lower_q)\n",
" prod_dict[i][\"upper_q_s\"].append(upper_q)\n",
" prod_dict[i][\"samples\"].append(samples)\n",
"\n",
"for i in prod_dict:\n",
" means_to_plot = np.array(prod_dict[i][\"means_to_plot\"]).squeeze()\n",
" lower_q_s = np.array(prod_dict[i][\"lower_q_s\"]).squeeze()\n",
" upper_q_s = np.array(prod_dict[i][\"upper_q_s\"]).squeeze()\n",
" samples = np.array(prod_dict[i][\"samples\"]).squeeze(-1)\n",
" y_true = np.array(y_train_prod[i][-DAYS_TO_PLOT:])\n",
"\n",
" model_names = [res[\"model\"].__class__.__name__ for res in results]\n",
"\n",
" print(\"mean shape\", means_to_plot.shape)\n",
" print(\"lower shape\", lower_q_s.shape)\n",
" print(\"upper shape\", upper_q_s.shape)\n",
" print(\"y_true shape\", y_true.shape)\n",
" print(\"samples shape\", samples.shape)\n",
"\n",
" #plot_predictions(y_true, means_to_plot, upper_q_s, lower_q_s, [res[\"model\"].__class__.__name__ for res in results])\n",
" plot_model_comparisons(samples, y_true, model_names=model_names)\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Log-likelihood"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Model Treeffuser has a NLL of {'nll_samples': 1.2175196664428274}\n",
"Model Treeffuser has a CRPS of {'crps_100': 0.6524468580538034}\n"
]
}
],
"source": [
"for res in results:\n",
" model = res[\"model\"]\n",
" name = res[\"model_name\"]\n",
" nll = LogLikelihoodFromSamplesMetric(n_samples=100).compute(model=model, X_test=X_test_np, y_test=y_test_np, samples=res[\"y_samples\"])\n",
" crps = CRPS().compute(model=model, X_test=X_test_np, y_test=y_test_np, samples=res[\"y_samples\"])\n",
" print(f\"Model {name} has a NLL of {nll}\")\n",
" print(f\"Model {name} has a CRPS of {crps}\")"
]
},
{
"cell_type": "code",
"execution_count": 39,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"{'mae': 0.9883351454782684, 'rmse': 2.1792834439725737, 'mdae': 0.5293582833519341, 'marpd': 141.8933048278295, 'r2': 0.6827293677776689, 'corr': 0.8279086805309621}\n"
]
},
{
"ename": "AssertionError",
"evalue": "",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mAssertionError\u001b[0m Traceback (most recent call last)",
"Cell \u001b[0;32mIn[39], line 3\u001b[0m\n\u001b[1;32m 1\u001b[0m accuracy \u001b[38;5;241m=\u001b[39m AccuracyMetric()\u001b[38;5;241m.\u001b[39mcompute(model\u001b[38;5;241m=\u001b[39mmodel, X_test\u001b[38;5;241m=\u001b[39mX_test_np, y_test\u001b[38;5;241m=\u001b[39my_test_np)\n\u001b[1;32m 2\u001b[0m \u001b[38;5;28mprint\u001b[39m(accuracy)\n\u001b[0;32m----> 3\u001b[0m \u001b[38;5;28;01massert\u001b[39;00m \u001b[38;5;28;01mFalse\u001b[39;00m\n",
"\u001b[0;31mAssertionError\u001b[0m: "
]
}
],
"source": [
"accuracy = AccuracyMetric().compute(model=model, X_test=X_test_np, y_test=y_test_np)\n",
"print(accuracy)\n",
"assert False"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Plot calibration plot"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def calibration_plot(y_samples: Float[np.ndarray, \"n_samples batch y_dim\"], y_test: Float[np.ndarray, \"batch y_dim\"]) -> None:\n",
" \"\"\"\n",
" We will plot the calibration plot for the model. Essentially, we will plot the\n",
" \"\"\"\n",
" assert y_test.shape[1] == 1, \"Only works for univariate outputs\"\n",
" n_samples = y_samples.shape[0]\n",
" y_samples = np.maximum(y_samples, 0.0)\n",
" y_samples = np.round(y_samples).astype(int)\n",
" y_test = y_test.astype(int)\n",
"\n",
" # Filter out the zeros\n",
" #non_zero_idx = y_test > 0\n",
" #y_test = y_test[non_zero_idx]\n",
" #y_samples = y_samples[:, non_zero_idx]\n",
"\n",
" y_test_expanded = y_test[np.newaxis, :].repeat(n_samples, axis=0)\n",
" prob_of_event = np.mean(y_samples <= y_test_expanded, axis=0)\n",
" prob_of_event_sorted = np.sort(prob_of_event.flatten())\n",
" return np.linspace(0, 1, len(prob_of_event)), prob_of_event_sorted\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"y_ex = None\n",
"for res in results:\n",
" samples = res[\"y_samples\"]\n",
" model = res[\"model\"]\n",
" x,y = calibration_plot(samples, y_test_np)\n",
" plt.plot(x, y, label=res[\"model_name\"])\n",
" y_ex = y\n",
"\n",
"\n",
"plt.plot([0, 1], [0, 1], linestyle=\"--\", color=\"black\")\n",
"\n",
"plt.xlabel(\"Predicted Probability\")\n",
"plt.ylabel(\"True Probability\")\n",
"plt.legend()\n",
"\n",
"plt.show()\n",
"\n",
"# plot the distribution of y\n",
"\n",
"sns.histplot(y_ex, bins=20)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Quantile prediction plot for the model\n",
"def make_quantile_plot(results):\n",
" quantiles = np.linspace(0.9, 0.999, 10)\n",
"\n",
" means = []\n",
" stds = []\n",
"\n",
" for res in results:\n",
" samples = res[\"y_samples\"] # shape [n_samples, batch, y_dim]\n",
" samples = np.maximum(samples, 0)\n",
" samples = np.round(samples)\n",
"\n",
" m, s = [], []\n",
"\n",
" for q in quantiles:\n",
" quantile_samples = np.quantile(samples, q, axis=1)\n",
" mean = np.mean(quantile_samples, axis=0)\n",
" std = np.std(quantile_samples, axis=0)\n",
" m.append(mean)\n",
" s.append(std)\n",
"\n",
" m = np.array(m).squeeze()\n",
" s = np.array(s).squeeze()\n",
" plt.plot(quantiles, m, label=res[\"model_name\"])\n",
" plt.fill_between(quantiles, m - s, m + s, alpha=0.3)\n",
"\n",
" means.append(m)\n",
" stds.append(s)\n",
"\n",
"\n",
" true_quantiles = []\n",
" for q in quantiles:\n",
" true_quantiles.append(np.quantile(y_test_np, q))\n",
"\n",
" plt.plot(quantiles, true_quantiles, label=\"True\")\n",
"\n",
" # set log scale on y axis\n",
"\n",
"\n",
" # set log scale on x axis\n",
" plt.xscale(\"log\")\n",
" plt.legend()\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"make_quantile_plot(results)"
]
},
{
"cell_type": "code",
"execution_count": 40,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"testbed.models.treeffuser.Treeffuser"
]
},
"execution_count": 40,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"model_cls"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"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": "3.9.19"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment