Advanced Prophet Usage

You can scan through fbprophet docs and find many options how to tweak your model. Some of that functionality is moved to initialization stage to be compatible with Sklearn API. We will showcase the parts that were moved to initialization, but you can also look for other model parameters that could help fine-tuning your model

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
plt.style.use('seaborn')
plt.rcParams['figure.figsize'] = [12, 6]

from hcrystalball.utils import generate_tsdata
X, y = generate_tsdata(n_dates=365*2)

from hcrystalball.wrappers import ProphetWrapper

ProphetWrapper?

Advanced Holidays

For holidays, we are able to define instead of single boolean attribute distribution around given day. We define lower_window, upper_window and prior_scales

extra_holidays = {
    'Christmas Day':{'lower_window': -2, 'upper_window':2, 'prior_scale': 20},
#     'Good Friday':{'lower_window': -1, 'upper_window':1, 'prior_scale': 30}
}

Unusual Seasonalities

extra_seasonalities = [
    {
        'name':'bi-weekly',
        'period': 14.,
        'fourier_order': 5,
        'prior_scale': 10.0,
        'mode': None
    },
    {
        'name':'bi-yearly',
        'period': 365*2.,
        'fourier_order': 5,
        'prior_scale': 5.0,
        'mode': None
    },
]

Exogenous Variables

from sklearn.pipeline import Pipeline
from hcrystalball.feature_extraction import HolidayTransformer

extra_regressors = ['trend_line']
X['trend_line'] = np.arange(len(X))

prophet = ProphetWrapper(
    name='prophet',
)
prophet_extra = ProphetWrapper(
    extra_holidays=extra_holidays,
    extra_seasonalities=extra_seasonalities,
    extra_regressors=extra_regressors,
    name='prophet_extra',
)

pipeline = Pipeline([
    ('holidays_de', HolidayTransformer(country_code = 'DE')),
    ('model', prophet)
])
pipeline_extra = Pipeline([
    ('holidays_de', HolidayTransformer(country_code = 'DE')),
    ('model', prophet_extra)
])

prds = (pipeline.fit(X[:-10], y[:-10])
         .predict(X[-10:])
         .merge(y, left_index=True, right_index=True, how='outer')
         .tail(50))

prds.plot(title=f"MAE:{(prds['target']-prds['prophet']).abs().mean()}");

prds_extra = (pipeline_extra.fit(X[:-10], y[:-10])
         .predict(X[-10:])
         .merge(y, left_index=True, right_index=True, how='outer')
         .tail(50))

prds_extra.plot(title=f"MAE:{(prds_extra['target']-prds_extra['prophet_extra']).abs().mean()}");

Compared to non-tweaked model, we are able to better catch the series dynamics, but don’t win against roughly average predictions

prds = (ProphetWrapper().fit(X[:-10], y[:-10])
                 .predict(X[-10:])
                 .merge(y, left_index=True, right_index=True, how='outer')
                 .tail(50)
)
prds.plot(title=f"MAE:{(prds['target']-prds['prophet']).abs().mean()}");

Full Prophet Output

If you need, you can also pass full_prophet_output and get rich predict output

(ProphetWrapper(full_prophet_output=True, conf_int=True)
     .fit(X[:-10], y[:-10])
     .predict(X[-10:])
)

	trend	prophet_lower	prophet_upper	trend_lower	trend_upper	additive_terms	additive_terms_lower	additive_terms_upper	weekly	weekly_lower	weekly_upper	multiplicative_terms	multiplicative_terms_lower	multiplicative_terms_upper	prophet
2018-12-22	6.164270	0.954172	11.432086	6.164270	6.164270	-0.021702	-0.021702	-0.021702	-0.021702	-0.021702	-0.021702	0.0	0.0	0.0	6.142568
2018-12-23	6.161881	1.040729	11.405363	6.161881	6.161881	-0.041156	-0.041156	-0.041156	-0.041156	-0.041156	-0.041156	0.0	0.0	0.0	6.120725
2018-12-24	6.159492	0.830015	11.012526	6.159492	6.159492	-0.028976	-0.028976	-0.028976	-0.028976	-0.028976	-0.028976	0.0	0.0	0.0	6.130516
2018-12-25	6.157103	1.019362	11.241077	6.157103	6.157103	0.007263	0.007263	0.007263	0.007263	0.007263	0.007263	0.0	0.0	0.0	6.164366
2018-12-26	6.154714	0.768980	11.470873	6.154714	6.154714	0.006438	0.006438	0.006438	0.006438	0.006438	0.006438	0.0	0.0	0.0	6.161152
2018-12-27	6.152325	0.794946	11.404954	6.152325	6.152325	0.058351	0.058351	0.058351	0.058351	0.058351	0.058351	0.0	0.0	0.0	6.210676
2018-12-28	6.149936	1.281950	12.100341	6.149902	6.149936	0.019782	0.019782	0.019782	0.019782	0.019782	0.019782	0.0	0.0	0.0	6.169718
2018-12-29	6.147547	0.613364	11.355996	6.147472	6.147573	-0.021702	-0.021702	-0.021702	-0.021702	-0.021702	-0.021702	0.0	0.0	0.0	6.125845
2018-12-30	6.145158	0.522798	11.447036	6.145007	6.145233	-0.041156	-0.041156	-0.041156	-0.041156	-0.041156	-0.041156	0.0	0.0	0.0	6.104002
2018-12-31	6.142769	0.611300	11.365543	6.142528	6.142896	-0.028976	-0.028976	-0.028976	-0.028976	-0.028976	-0.028976	0.0	0.0	0.0	6.113793