# Multi-Season Overlays

# Overview

• Each time series can be de-composed into multiple components consisting of a trend, one or multiple seasonal cycles, and the intrinsic residuals often called noise.
• The seasonality can be deduced manually by observing the records using graphs or determined automatically by analyzing periodograms.
• Once the seasonality is known, overlaying multiple time intervals with the same seasonal lag can be used as an essential visual analysis technique.
• The overlaid series can be assigned different weights by applying opacity to more distant seasons.

# Keywords

overlay, lag, shift, seasonality

# Graphics

The chart shows the Oroville Dam water levels for the current year (2018, red series) with prior year overlays displayed with a gradually decreasing visibility. Year 2009 is the least visible.

# Syntax Features

• time-offset setting to add time lag to the underlying series.

[series]
  time-offset = @{offset} year
  # time-offset = 1 year

• for loop to iterate over multiple seasons.

for name in offsets
  # access current element by @{name}
endfor

• range function to generate a numeric sequence, for example a range of years to shift the series.

# inline array
for offset in range(1,10)