# Statistical Advisor, Explore/Summarize a Time Series

Use TIME SERIES. This chapter describes varous techniques for displaying as well as transforming/smoothing time series data.

Time series data typically arise when one studies one or more variables over time; for example, one may track the price of stock over successive trading days. Some time series software programs contain a modeling/transformation 'workbench' which allows you to display or plot the series of stock quotes. You can then choose from among many different transformations to smooth the data, so as to remove random 'noise' fluctuations in the stock price. When reviewing plots of such smoothed data, trends or recurring patterns over time often become more clearly visible.

Additional advanced time series techniques described in this chapter include autocorrelation analysis, Fourier decomposition, and seasonal and non-seasonal ARIMA. The purpose of all of these methods is to detect patterns or structure in the movement of the respective variable(s) over time.