Identify if single cell contains multiple substrings

[Jay Hornberg]

Hi,

I'd like to identify observations when a string variable contains multiple substrings.

In my dataset, I have a string variable containing crime descriptions. For that string variable, I want to identify when "REG" and "GUN" appear together in the same cell.

The data is messy and not standardized. Here is an example of how some of the strings look:

crime
GUN OFFENDER REGISTRATION
GUN OFFENDER-FAIL TO REGISTER
GUN OFFENDER/FAIL REG OFFENDER
GAS/AIR/PAINTBALL GUN: POSSESS
FIRING HANDGUN IN CITY LIMITS
FRAUDULENT POSSESSION OF VEH OWNERSHIP REG. PLATE
KNOWINGLY HOLDING FALSIFIED VEH. REG. PLATE

I've successfully used the strpos command to isolate observations containing a single substring, i.e. :

l if strpos(crime, "REG")
l if strpos(crime, "GUN")

And I've been able to identify observations that contain either one substring or another, i.e. :

l if strpos(crime, "REG" "GUN")

But I haven't been able to figure out how to identify if a single cell contains both "REG" and "GUN".

Any advice is appreciated.

[Nick Cox]

l if strpos(crime, "REG") & strpos(crime, "GUN")

[Mike Lacy]

(crossed in the ether, deleted.)

[Nick Cox]

I am surprised to hear that

l if strpos(crime, "REG" "GUN")

works and guess that's an illusion.

Consider this:

. list

     +---------+
     |    var1 |
     |---------|
  1. | REG GUN |
  2. |     REG |
  3. |    frog |
  4. |    toad |
     +---------+

. list if strpos(var1, "REG" "GUN")

     +---------+
     |    var1 |
     |---------|
  1. | REG GUN |
  2. |     REG |
     +---------+

I think Stata is just ignoring "GUN" there.

[Jay Hornberg]

Thank you very much, Nick!