So far, we have discussed how MA can be used for estimating the trend-cycle and seasonal components of a time series. The method of MA works under the simple assumption that seasonal changes are constant over consecutive years, weeks, or a period suitable for the given use case. However, constant seasonality might be valid for several applications that require advanced method such as Seasonality and Trend decomposition using Locally Weighted Smoothing of Scatter plot also commonly referred as the STL method.
In this section, we will tackle time series with complex patterns using the Python statsmodels.tsa package. Our objective would be to estimate the trend-cycle and seasonal components. Besides, ...
