Collapsing different variables in one variable

[Vincenz Jahn]

I have different variables, resulting from generating new variables with regexs and regexm. Since these numbers I was looking for have different formats, I created redundant values.

Ref3	Ref4	Ref5	Ref6
Ref. 2589	Ref. 258
Ref. 6265	Ref. 626
Ref. 3403	Ref. 340	Ref. 34

My goal is to create a new variable where only the longest string or the longest number in one row appears. Any help kindly appreciated!

RefNew
Ref. 2589
Ref. 6265
Ref. 3403

[David Dauria]

There's probably a half dozen ways to do this. My first crack at it would be

gen RefNew =""
forvalues i = 3/6 {
replace RefNew = Ref`i' if strlen(Ref`i')>strlen(Ref`i')
  }

I am assuming you have a larger dataset than the one you presented, so it's not always the case that Ref3 is the longest variable.

[William Lisowski]

I think David meant

  replace RefNew = Ref`i' if strlen(Ref`i')>strlen(RefNew)

and it can be made slightly tighter with

gen RefNew = Ref3
forvalues i = 4/6 {
  replace RefNew = Ref`i' if strlen(Ref`i')>strlen(RefNew)
  }

[William Lisowski]

Also posted at http://www.statalist.org/forums/foru...y-of-variables with a different answer given.

[Vincenz Jahn]

Thank you for your help! Works perfectly