Similar names in a variable list

[Joseph Chav]

Statalist,

I have a list of 2000 drugs; among those, there are drugs with slightly different names/doze. I would like to combine similar names. For example, below, each color would be considered a drug (class). Is there any way to do it rather than checking one by one, then recoding in STATA?
Thank you

[Daniel Shin]

You may need to do some manual checking, but you could start with just taking the first word or two as the name of the active ingredient.

Using the first word only:

gen class=word(lower(drug),1)

If you have multiple active ingredients, you will need to parse them out.