Dropping variables in the batch structure

[Sonnen Blume]

My dataset contains ~4000 variables from a survey that collected data on upto 10 members from the same household. (AGE_01, AGE_02...AGE_10/SEX_01, SEX_02.... SEX_10/BMI_01, BMI_02.....BMI_10).

Is it possible to keep only the FIRST observation for all the variables (AGE1, SEX1, BMI1) by dropping the rest.

[Mike Lacy]

Something in the neighborhood of:

keep *_01 x y z

where x y z represent any variables you want besides the ones referenced by *_01, should work. See -help varlist- and -help keep-.

[Sonnen Blume]

Something in the neighborhood of:

keep *_01 x y z

where x y z represent any variables you want besides the ones referenced by *_01, should work. See -help varlist- and -help keep-.

Thanks very much Mike. keep *_01 did the trick. This method retains all variables containing "_01", which is little troublesome as some of the variables are not numbered and still need to be retained. I didn't foresee this actually. Is there a way to drop solely the variables numbered as:
_02, _03....

[Rich Goldstein]

try

drop *_02 *_03 *_04 ...

[Sonnen Blume]

try

drop *_02 *_03 *_04 ...

It worked. Thanks so much!