Tell me more ×
Cross Validated is a question and answer site for statisticians, data analysts, data miners and data visualization experts. It's 100% free, no registration required.
up vote 0 down vote favorite

How should I represent a lack of information in my lag locations when performing time series prediction with an explanatory set of multiple variables and a response set of 1 variable, with a 0 or NaN? What is preferred and for what reason, what impacts on my network prediciton (output) will I see if I choose either 0 or NaN?

0.789389068 NaN         NaN
0.635048232 0.789389068 NaN
0.602893891 0.635048232 0.789389068
0.459807074 0.602893891 0.635048232
0.752411576 0.459807074 0.602893891
0.718649518 0.752411576 0.459807074

OR

0.789389068 0           0
0.635048232 0.789389068 0
0.602893891 0.635048232 0.789389068
0.459807074 0.602893891 0.635048232
0.752411576 0.459807074 0.602893891
0.718649518 0.752411576 0.459807074
share|improve this question
If your software can handle it, probably NaN is preferred. This will make it deal with the missing information in a reasonable way. Otherwise you may want to use the average or some other 'smart' choice of values rather than zero, but that depends on your model ofcourse. – Dennis Jaheruddin Jan 18 at 10:16

Know someone who can answer? Share a link to this question via email, Google+, Twitter, or Facebook.

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.