Literature Series
Time series analysis – missing data points at start and end time points?
I’m doing sime series analysis on a set of data from 20 study sites. Some of these sites have missing data for start and end time points (eight time points in total). I’m doing two-point moving averages to analyse trends over time but I’m not sure how to cope with missing data when I need all the calculations I can get!
If I have missing T0 and T1 data but T2 and T3 are present, how do I do a valid T1-T2 moving average or do I just use T2 data alone without averaging and then use a moving average for T2-T3?
I’d appreciate any references to literature to use as evidence regarding my strategy to cope with my missing data too, if possible?
I wouldn’t recommend correcting for missing data points as these are endpoints. The reason is that you have to extrapolate, which will introduce a much big error than in the case of missing interior points.
If your goal is to model some variables as function of the others, then a good method is to add the missing data points as parameters in the model.
Classic Literature Series – Charles Dickens
Tags: books, literature, literature series, literature series blue, literature series story, literature series textbooks, reading, reference, reviews, riverside literature series
This entry was posted on Friday, March 20th, 2009 at 4:23 am and is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.