Condense data

[Matt Sveum]

I have a panel dataset with an unequal number of observations per company. In other words, some companies might show up in the data five times and others might show up 11 times. I have 11 years total. Here is a sample of what I have:

Company rank13 rank 12 rank11 rank10
A 1 . . .
A . 1 . .
B . . . 2

I would like to create one observation for each company with all ranks. For example:

Company rank13 rank12 rank11 rank10
A 1 2 . .
B . . . 2

I have tried to do this with the -reshape- command, but I can't get it to do what I want. I have also searched Google and Statalist, but haven't been able to find anything that applies.

Thank you advance for your help.

[Richard Williams]

I feel like some key info is missing. Is there a year variable? Can you only have a value once on a rank variable and in every other year you have to be missing on every rank variable? Maybe this is really a collapse problem, where you take the first/ only non missing value for each variable for each case.

[Matt Sveum]

The -collapse- command was exactly what I was looking for. It worked perfectly. Thank you for your help.

[Richard Williams]

Great. In your example, after you collapse to one observation per company, I don't know why you would expect the 2 value for company A. Was that a mistake or am I missing something?

[Matt Sveum]

Ahh, yes, that was a typo. It should have been a 1. Sorry about that.