mlforecast 0.3.0

Scalable machine learning based time series forecasting

mlforecast

Install

PyPI

pip install mlforecast

If you want to perform distributed training, you can instead use pip install mlforecast[distributed], which will also install dask. Note that you’ll also need to install either LightGBM or XGBoost.

conda-forge

conda install -c conda-forge mlforecast

Note that this installation comes with the required dependencies for the local interface. If you want to perform distributed training, you must install dask (conda install -c conda-forge dask) and either LightGBM or XGBoost.

How to use

The following provides a very basic overview, for a more detailed description see the documentation.

Store your time series in a pandas dataframe in long format, that is, each row represents an observation for a specific serie and timestamp.

from mlforecast.utils import generate_daily_series

series = generate_daily_series(
    n_series=20,
    max_length=100,
    n_static_features=1,
    static_as_categorical=False,
    with_trend=True
)
series.head()

	ds	y	static_0
unique_id
id_00	2000-01-01	1.751917	72
id_00	2000-01-02	9.196715	72
id_00	2000-01-03	18.577788	72
id_00	2000-01-04	24.520646	72
id_00	2000-01-05	33.418028	72

Next define your models. If you want to use the local interface this can be any regressor that follows the scikit-learn API. For distributed training there are LGBMForecast and XGBForecast.

import lightgbm as lgb
import xgboost as xgb
from sklearn.ensemble import RandomForestRegressor

models = [
    lgb.LGBMRegressor(),
    xgb.XGBRegressor(),
    RandomForestRegressor(random_state=0),
]

Now instantiate a Forecast object with the models and the features that you want to use. The features can be lags, transformations on the lags and date features. The lag transformations are defined as numba jitted functions that transform an array, if they have additional arguments you supply a tuple (transform_func, arg1, arg2, …).

from mlforecast import Forecast
from window_ops.expanding import expanding_mean
from window_ops.rolling import rolling_mean

fcst = Forecast(
    models=models,
    freq='D',
    lags=[7, 14],
    lag_transforms={
        1: [expanding_mean],
        7: [(rolling_mean, 7)]
    },
    date_features=['dayofweek'],
    differences=[1],
)

To compute the features and train the models call fit on your Forecast object. Here you have to specify the columns that:

• Identify each serie (id_col). If the series identifier is the index you can specify id_col='index'
• Contain the timestamps (time_col). Can also be integers if your data doesn’t have timestamps.
• Are the series values (target_col)

fcst.fit(series, id_col='index', time_col='ds', target_col='y', static_features=['static_0'])

Forecast(models=[LGBMRegressor, XGBRegressor, RandomForestRegressor], freq=<Day>, lag_features=['lag-7', 'lag-14', 'expanding_mean_lag-1', 'rolling_mean_lag-7_window_size-7'], date_features=['dayofweek'], num_threads=1)

To get the forecasts for the next 14 days call predict(14) on the forecast object. This will automatically handle the updates required by the features using a recursive strategy.

predictions = fcst.predict(14)

import matplotlib.pyplot as plt
import pandas as pd

fig, ax = plt.subplots(nrows=2, ncols=2, figsize=(12, 6), gridspec_kw=dict(hspace=0.3))
for i, (cat, axi) in enumerate(zip(series.index.categories, ax.flat)):
    pd.concat([series.loc[cat, ['ds', 'y']], predictions.loc[cat]]).set_index('ds').plot(ax=axi)
    axi.set(title=cat, xlabel=None)
    if i % 2 == 0:
        axi.legend().remove()
    else:
        axi.legend(bbox_to_anchor=(1.01, 1.0))
fig.savefig('figs/index.png', bbox_inches='tight')
plt.close()