Sometimes the data that we will analyze is a variable measured at fixed time intervals; when we have such data, we are talking about a time series. More specifically, at each step of the time series, there is more than one possible outcome and part of the outcome for each step is randomized and might only depend on a few steps back in time. For these reasons, simple linear regression does not work. In time series analysis, we build models to explain the variations in time, which is sometimes referred to as longitudinal analysis.
This chapter covers the following topics in time series analysis:
Time series modeling, its usefulness, and how Pandas handles data
Various common patterns in time series
The concept of stationarity and how to test and make your data stationary
Resampling, smoothing, and calculating rolling statistics
How to model the known variations and make short forecasts
We start off with some more information about time series and what insights analyzing...