We have seen that doing a preliminary data analysis and visualizing the dataset is the first step in any machine learning project. Time series are no exception. So, we will start by exploring time series and learning about its different characteristics.
In the case of a time series, a preliminary analysis implies modeling it; that is, understanding whether it is periodic, whether it shows a given tendency (increasing or decreasing with time), or whether it is stationary (mean and variance of the values don't change over time), among other measures. Visualization plays a fundamental role in this analysis, since many of the time series characteristics can be deduced using a graphical representation of the data points, even if there are numerical methods to calculate them.
Let's use a popular dataset to illustrate the modeling and visualization...
