Categorical variables Automatic recoding to start with "1" and continues sequentially

[Riham Omar]

Hi everyone, I need to recode categorical variables so that its values are recoded automatically in sequential order starting by "1"
for example now I have a variable school_lev where its values are 10"None" 33 "elementary school" 40"high school" so i want to somehow generate the variable so that 10>> becomes 1 33>>becomes 2 & 40>>become 3. of course I know Ican do it manually but with the number of variables i have, it will take very long time to do it . Plus if there is a way to preserve the original labels please let me know.
Many thanks in advance

[Rich Goldstein]

the main part of what you want can be done via

ta oldvar, gen(newvar)

replace "oldvar" and "newvar" with your actual variable name and the name you want for the new variable

[Nick Cox]

Not so; tabulate just generates dummy variables with that syntax.

egen newvar = group(oldvar), label

is my suggestion.

[Riham Omar]

the main part of what you want can be done via

ta oldvar, gen(newvar)

replace "oldvar" and "newvar" with your actual variable name and the name you want for the new variable

Thank you so much for your reply but as far as I understand, using this way we are creating dummy variables that equal the number of values we have. while I need to have only one variable with the new sequential codes

[Riham Omar]

Not so; tabulate just generates dummy variables with that syntax.

egen newvar = group(oldvar), label

is my suggestion.

Thank you soooo much. this actually worked. However, I will need to remove the value before the value labels as it is preserved as well in the process. as now for example 1 [10]:none

[Nick Cox]

That means applying numlabel first.

[Sorrel Namaste]

Thank you Riham for the question and for your helpful response Nick. This posting is similar to something I am trying to do but could use some additional guidance. I was able to recode automatically and sequentially using the code provided (similar to Riham's issue but my oldvar and newvar are numeric). However, my dataset is a merged dataset of 100 surveys which are identified as 1 through 100. I would like to apply the egen to each of these surveys. I created a forvalues to loop through each of the 100 survey (variable name called survey in the example below). It codes correctly for the first survey but then will not move on because the variable has already been generated. Any suggestions on how to either modify the current code or another way to accomplish the same thing?

forvalues i =1/100{
egen newvar = group(oldvar)if group==`i' } variable newvar already defined