Exponential smoother/forecaster with de-seasonalization
Smooth models an unobserved level and trend component in a noisy signal, along with optional "seasonal" effects at the level of day-over-day variation. The result is a smoothed version of the signal which may be used to forecast future values or detect unexpected variation.
In this example, we apply our smoother to a sample 10-day timeseries published in Twitter's AnomalyDetection package. The series is 10 days of counts, with a regular daily pattern and occasional surprises (spikes and dips). The Juttle output focuses on the three days around Oct 1.