Hands-On Data Analysis with Pandas

Time series decomposition

As mentioned in Chapter 1, Introduction to Data Analysis, time series can be decomposed into trend, seasonal, and remainder components utilizing a specified frequency. This can be achieved with the statsmodels package, which StockModeler.decompose() is using:

>>> from stock_analysis import StockModeler>>> decomposition = StockModeler.decompose(nflx, 20)>>> fig = decomposition.plot()>>> fig.suptitle('Netflix Stock Price Time Series Decomposition', y=1)

This returns the decomposition plot for Netflix with a frequency of 20 trading days:

For more complicated models, we could decompose and then build our model around ...