9 Time series data: Analysis
This chapter covers
- Analyzing time series data to answer a research question
- Uncovering hidden depth in seemingly simple time series
- Evaluating whether a time series dataset is appropriate for forecasting
- Examining the constituent parts of a time series
- Building a forecasting model
In this chapter, we continue exploring the value of time series data. In chapter 8, we explored raw time series data and decided which records to keep before analyzing the time series further. This chapter is about the second part of the process: analyzing time series data to look for patterns, as well as decomposition and forecasting. We will first recap the project brief from the previous chapter and summarize the work done so far before continuing with the analysis.
9.1 Project 6 revisited: Analyzing time series to improve cycling infrastructure
We prepared our data, and it is ready for analysis. But before we start the analysis, let’s recap the problem statement and the data dictionary from the previous chapter. The data is available for you to attempt the project yourself at https://davidasboth.com/book-code. You will find the files that you can use for the project, as well as the example solution in the form of a Jupyter notebook. The notebook for this chapter picks up where chapter 8 left off.