Ipolate & Epolate

[Anupa Sahdev]

Hello,

I have a question regarding the ipolate command in STATA. I have missing data for the first year of data for a variable and believe that the variable is a function of time.
I would like to know how stata has calculated the imputed values for the missing data when I use a command with the following structure:

by ID: impolate Y Year, gen(iY) epolate

In particular how has it used the available data for any given panel.

Many thanks

A

[Nick Cox]

The ipolate command uses linear interpolation, including extrapolation when requested. This is exactly what, I guess, you did as a small child when you drew straight lines connecting points to see what emerged from a puzzle picture. It is documented in the help and the manual entry and also at e.g. http://en.wikipedia.org/wiki/Linear_interpolation

Doing this separately by panels is precisely that: the interpolation and extrapolation for any given panel is entirely separate in the sense that no information is used from any other panel.

As is often true, the best way to see what interpolation did is to graph the results.

[Adithya Kolloju]

Hi Nick,

I was wondering if to use the command ipolate and mipolate command in Stata. I was using my data set about universities with 27 variables from period 2001 - 2020. and expectedly I have some missing values with the variables. so, I was using ipolate command along with epolate at the end. But when I browse the data to check the differences. But I see a lot of differences and each observation aren't same ( I mean there is no similarity with the existed and new values generated). Could you help me with thsi? Thank you, Nick.

[Nick Cox]

#3 Please post example data and exact commands used to make your problems precise.