Chapter 4: Interpolation Using PROC EXPAND
4.1 Interpolation of Time Series
4.1 Interpolation of Time Series
Most SAS procedures require that the time series be observations equidistant in time. This means that a problem appears if one observation is missing, such as when no information is available for a month. For a series of total sales of a particular item, no sales information within a month in principle means no sales at all that month. But such missing observations could be due to errors in the accounting systems. If so, no registrations does not mean a zero value. The price of a share at a stock market is typically recorded every time the share is traded, but in theory, it could also be traded at other points in time ...
