Creating dummy variables

[Alexandra Sotiriou]

Dear Statalists,

This is my first post.

I have a challenge in generating dummy variables using a variable list which contains non-numeric values (so labels of country codes). Specifically I have a long list of country codes (i.e. AT, BE ext) and I want to create 4 dummies corresponding to three blocks of countries i.e: EU_North (if AT, BE DE) EU_South (if PT, SP) ext and I type the following two commands and get the following message:

generate EU_total = 0

1. replace EU_total = 1 if countrycode == AT | BE | FR | DE | GB | IE | ES | IT | NL | LU | PT | SE | FI
AT not found
r(111);

2. replace EU_total = 1 if countrycode == AT | countrycode == BE | countrycode == FR | countrycode == DE | countryc
> ode == GB | countrycode == IE | countrycode == ES | countrycode == IT | countrycode == NL | countrycode ==
> LU | countrycode == PT | countrycode == SE | countrycode == FI
AT not found
r(111);



I have converted countrycode from string to non numeric

describe countrycode

storage display value
variable name type format label variable label
------------------------------------------------------------------------------------------------------------
countrycode long %8.0g countrycode
ΚΩΔ_ΧΩΡ


Does this mean that I cannot use labels in the dummy command?

I then want to collapse (sum) my values based on the 4 dummies to get aggregate figures for each dummy/country block.

Thank you for any ideas or insight you might have!