Page 1 of 1

Data manipulation

Posted: Tue Aug 03, 2010 11:27 am
by y.khairulin
Hi,
I am new to eviews, so please be patient for my simple and may be obvious questions.
I have data in the following format (see attachment), I imported data in unbalanced/undated format to eviews. As you can see gvkey and year series don't correspond to unique opt_bs observations.
What I need is to generate a new series that will take an arithmetic/weighed average of the observation values corresponding to company specific gvkey and year.
Basically the new series should generate for example: gvkey 001004, year 1997, opt_bs = (1076.006+1452.617)/2.
In other words, there are several option grants to company CEO during the year, but I need an average value of the option grant for a given year, so that later this data can be
structured as dated panel and used in my regression. Thank you in advance for your help.

Re: Data manipulation

Posted: Wed Aug 11, 2010 11:14 am
by EViews Glenn
Structure your panel as undated with identifiers GVKEY and YEAR. EViews will create a new crossid variable indexing the different OPT_BS responses. Then @meansby will create a new series with the values you want. Restructure again as desired.

Re: Data manipulation

Posted: Wed Aug 11, 2010 6:19 pm
by y.khairulin
Thanks a lot, that helped.