Unbalanced Panel Manipulation

[theodore04]

Dear All,

I have an unbalanced panel and want/wish to make it balanced. Is there a quick way of doing this?

Is there a way of replacing the NA values with the average of the observation before and after or something like this?

regards

[EViews Gareth]

You can use interpolation (from the Proc menu on a series) to fill in the NAs like that. Its just simple linear interpolation.

[theodore04]

thanks.

is there any problem when one has a panel dataset?

what happens when for a cross section the first obs is missing?

is there a danger that is using info from the previous cross section?

thanks

[EViews Gareth]

No, it won't use a value from a previous cross-section. It will remain as an NA (since interpolation requires observations before and after the missing observation, if the first value for a cross-section is missing, it cannot interpolate).