Sum across many columns or many rows?

[Peter Meier]

Dear Statalisters,

I have a dataset with variables var_1:var_200 as well as a variable id that takes values 1:200 as well, so that the total dataset has 200 rows and 201 columns including the "id" variable.

Now I would like to sum first across all rows, and secondly across all but the first column (I could drop the first if necessary). How can I do that?

I've tried summing across the columns with "egen example = rowtotal()", but would need to type each of the 200 variables there? Or can I at least tell Stata more succintly to sum all of var_1 through var_200?

Thank you so much and best regards,
PM

[Carlo Lazzaro]

Peter:
if your variabes are actually named -var_1-, -var_2- and so on, you can use the wildcard symbol (*) to speed things up:

. set obs 10
number of observations (_N) was 0, now 10

. g var_1=runiform()

. g var_2=runiform()

. egen sum=rowtotal(var_*)

. list

     +--------------------------------+
     |    var_1      var_2        sum |
     |--------------------------------|
  1. | .3488717   .2047095   .5535812 |
  2. | .2668857   .8927587   1.159644 |
  3. | .1366463   .5844658   .7211121 |
  4. | .0285569   .3697791    .398336 |
  5. | .8689333   .8506309   1.719564 |
     |--------------------------------|
  6. | .3508549   .3913819   .7422367 |
  7. | .0711051   .1196613   .1907664 |
  8. |  .323368   .7542434   1.077611 |
  9. | .5551032   .6950234   1.250127 |
 10. |  .875991   .6866152   1.562606 |
     +--------------------------------+

.

[Romalpa Akzo]

1. What Peter want is:

egen example1=rowtotal(var_1-var_200)

Note the dash(-). Of course, you have to check and type exactly the name of the second variable (after id, i.e var_1) and the last (i.e var_200).

2. The more convenient code might be:

ds id, not
egen example2=rowtotal(`r(varlist)')

which requires the name of the very first variable (i.e id) only.

[Peter Meier]

Thank you Carlo and Romalpha!

Both the version with the wildcard and the one with the dash are working now!

All best, PM