Seasonality and Seasonal Decomposition Analysis of hMPV
We evaluated the number and incidence of weekly hMPV cases and
hMPV-associated ALRI cases from 2011–2016. We used
R’s23 (R Version 4.0.4) decompose (part of the stats
package) and forecast package 24 to assess the
temporal dynamics of hMPV cases and specifically isolate the trend,
seasonal, and error components. Since the magnitude of the seasonal
fluctuations and the variation around the trend-cycle do not vary
proportionally with time, we used an additive time-series decomposition
approach to isolate the temporal trend, seasonality, and error
components. In additive decomposition we assume:
\(y\)t = St + Tt +
Rt
Where \(y\)t is the data, St is
the seasonal component, Tt is the trend
component, and Rt is the remainder component.