Today we launch the second tutorial in our ongoing series about data preparation in time series analysis. Our focus this time around is one of the most fundamental assumptions in time series analysis: Stationarity, or the assumption that a data sample is drawn from a stationary process.
First, we lay down a bedrock definition of stationarity and demonstrate the minimum requirements for stationarity in our time series analysis. Then we’ll tackle a time series data sample – namely the closing prices for IBM stock from January 3rd, 2012 to the present date – which we’ll use to develop some observations about stationary processes, and explore the underlying intuitions behind them.
You can find more details and a step-by-step tutorial, along with a downloadable spreadsheet and PDF, at the link below:
Data preparation no. 2: Stationarity
Questions? Comments? We’d love to hear your ideas for future issues or feedback on past issues. Contact us at support@numxl.com.
