K
Kenbwire / Jun 27 2019

Multiplicative Seasonality

#%load_ext rpy2.ipython 
#%matplotlib inline
from fbprophet import Prophet
import pandas as pd
import numpy as np
from matplotlib import pyplot as plt
import logging
logging.getLogger('fbprophet').setLevel(logging.ERROR)
import warnings
warnings.filterwarnings("ignore")
Shift+Enter to run
Python
Python 3
%%R
library(prophet)
Shift+Enter to run
Python
Python 3

By default Prophet fits additive seasonalities, meaning the effect of the seasonality is added to the trend to get the forecast. This time series of the number of air passengers is an example of when additive seasonality does not work:

example_air_passengers.csv
%%R -w 10 -h 6 -u in
df <- read.csv('../examples/example_air_passengers.csv')
m <- prophet(df)
future <- make_future_dataframe(m, 50, freq = 'm')
forecast <- predict(m, future)
plot(m, forecast)
Shift+Enter to run
Python
Python 3
nil
df = pd.read_csv(
example_air_passengers.csv
)
m = Prophet()
m.fit(df)
future = m.make_future_dataframe(50, freq='MS')
forecast = m.predict(future)
fig = m.plot(forecast)
Shift+Enter to run
Python
Python 3
nil

This time series has a clear yearly cycle, but the seasonality in the forecast is too large at the start of the time series and too small at the end. In this time series, the seasonality is not a constant additive factor as assumed by Prophet, rather it grows with the trend. This is multiplicative seasonality.

Prophet can model multiplicative seasonality by setting seasonality_mode='multiplicative' in the input arguments:

%%R -w 10 -h 6 -u in
m <- prophet(df, seasonality.mode = 'multiplicative')
forecast <- predict(m, future)
plot(m, forecast)
Shift+Enter to run
Python
Python 3
nil
m = Prophet(seasonality_mode='multiplicative')
m.fit(df)
forecast = m.predict(future)
fig = m.plot(forecast)
Shift+Enter to run
Python
Python 3
nil

The components figure will now show the seasonality as a percent of the trend:

%%R -w 9 -h 6 -u in
prophet_plot_components(m, forecast)
Shift+Enter to run
Python
Python 3
fig = m.plot_components(forecast)
Shift+Enter to run
Python
Python 3
nil

With seasonality_mode='multiplicative', holiday effects will also be modeled as multiplicative. Any added seasonalities or extra regressors will by default use whatever seasonality_mode is set to, but can be overriden by specifying mode='additive' or mode='multiplicative' as an argument when adding the seasonality or regressor.

For example, this block sets the built-in seasonalities to multiplicative, but includes an additive quarterly seasonality and an additive regressor:

%%R
m <- prophet(seasonality.mode = 'multiplicative')
m <- add_seasonality(m, 'quarterly', period = 91.25, fourier.order = 8, mode = 'additive')
m <- add_regressor(m, 'regressor', mode = 'additive')
Shift+Enter to run
Python
Python 3
m = Prophet(seasonality_mode='multiplicative')
m.add_seasonality('quarterly', period=91.25, fourier_order=8, mode='additive')
m.add_regressor('regressor', mode='additive')
Shift+Enter to run
Python
Python 3
nil

Additive and multiplicative extra regressors will show up in separate panels on the components plot.

conda update conda
pip install pystan
pip install fbprophet
184.0s
Bash in Python
Python 3