API Reference
Feature engineering
Tutorials
How to Guides
Experiments
Model References
- ADIDA Model
- ARCH Model
- ARIMA Model
- AutoARIMA Model
- AutoCES Model
- AutoETS Model
- AutoRegressive Model
- AutoTheta Model
- CrostonClassic Model
- CrostonOptimized Model
- CrostonSBA Model
- Dynamic Optimized Theta Model
- Dynamic Standard Theta Model
- GARCH Model
- Holt Model
- Holt Winters Model
- IMAPA Model
- MFLES
- Multiple Seasonal Trend (MSTL)
- Optimized Theta Model
- Seasonal Exponential Smoothing Model
- Seasonal Exponential Smoothing Optimized Model
- Simple Exponential Smoothing Optimized Model
- Simple Exponential Smoothing Model
- Standard Theta Model
- TSB Model
API Reference
Feature engineering
Generate features for downstream models
source
mstl_decomposition
mstl_decomposition (df:Union[pandas.core.frame.DataFrame,polars.datafram e.frame.DataFrame], model:statsforecast.models.MSTL, freq:str, h:int)
Decompose the series into trend and seasonal using the MSTL model.
| Type | Details | |
|---|---|---|
| df | Union | DataFrame with columns [unique_id, ds, y]. |
| model | MSTL | Model to use for the decomposition. |
| freq | str | Frequency of the data (pandas alias) |
| h | int | Forecast horizon. |
| Returns | Tuple | Original dataframe with the ‘trend’ and ‘seasonal’ columns added. |
from fastcore.test import test_fail
from utilsforecast.losses import smape
from statsforecast.models import Naive
from statsforecast.utils import generate_series
series = generate_series(10, freq='D')
series['unique_id'] = series['unique_id'].astype('int64')
test_fail(lambda: mstl_decomposition(series, Naive(), 'D', 14), contains='must be an MSTL instance')
horizon = 14
model = MSTL(season_length=7)
series = series.sample(frac=1.0)
train_df, X_df = mstl_decomposition(series, model, 'D', horizon)
series_pl = generate_series(10, freq='D', engine='polars')
series_pl = series_pl.with_columns(unique_id=pl.col('unique_id').cast(pl.Int64))
train_df_pl, X_df_pl = mstl_decomposition(series_pl, model, '1d', horizon)
pd.testing.assert_series_equal(
train_df.groupby('unique_id')['ds'].max() + pd.offsets.Day(),
X_df.groupby('unique_id')['ds'].min()
)
assert X_df.shape[0] == train_df['unique_id'].nunique() * horizon
pd.testing.assert_frame_equal(train_df, train_df_pl.to_pandas())
pd.testing.assert_frame_equal(X_df, X_df_pl.to_pandas())
with_estimate = train_df_pl.with_columns(estimate=pl.col('trend') + pl.col('seasonal'))
assert smape(with_estimate, models=['estimate'])['estimate'].mean() < 0.1
model = MSTL(season_length=[7, 28])
train_df, X_df = mstl_decomposition(series, model, 'D', horizon)
assert train_df.columns.intersection(X_df.columns).tolist() == ['unique_id', 'ds', 'trend', 'seasonal7', 'seasonal28']
On this page