Stata Code

[Milena Schmidt]

Hey Stata Community,


I'm looking for a way to generate a variable depending on the "4 last letters" of another variable.


Example:

My variable has car names like "VW" or "AUDI", but sometimes the variable also says "VW FGDT" (4 random letters behind the car name, no model specifications).

How can I generate a new variable that is "1" if there is this random 4 letter code behind the car name? Is there something like a "placeholder" I can use for these random 4 letters?

Best regards,
Milena

[Carlo Lazzaro]

Milena:
do you mean something along the following lines?

. set obs 2
number of observations (_N) was 1, now 2

. g car_name="AUDI" in 1

. replace car_name="VW FGDT" in 2
variable car_name was str4 now str7


. split car_name
variables created as string:
car_name1  car_name2

. g wanted=1 if car_name1!="" & car_name2!=""

. replace wanted=0 if car_name1!="" & car_name2==""

. list

    +-----------------------------------------+
     | car_name   car_na~1   car_na~2   wanted |
     |-----------------------------------------|
  1. |     AUDI       AUDI                   0 |
  2. |  VW FGDT         VW       FGDT        1 |
     +-----------------------------------------+

.

.