Separating a character and a number part of a variable

[kamalesh pahurkar]

I have a variable consisting of characters and numbers in a string format. How can I separate the character part of a variable and the numbers part? The positioning of a number part is not in the same sequence for all observations so that I couldn't use a substring option. Is there any general way to identify the number part and separate it?

[Nick Cox]

Without a data example this is rather vague: do the numbers include fractional parts, for example? Here is a starter, using moss from SSC.

clear 
input str42 whatever 
"Answer is 42"
"101 Dalmatians"
"76 trombones"
"Stata 18: the mystery"
end

moss whatever , match("([0-9]+)") regex 

gen remainder = subinstr(whatever, _match1, "", .)

list 

     +------------------------------------------------------------------------+
     |              whatever   _count   _match1   _pos1             remainder |
     |------------------------------------------------------------------------|
  1. |          Answer is 42        1        42      11            Answer is  |
  2. |        101 Dalmatians        1       101       1            Dalmatians |
  3. |          76 trombones        1        76       1             trombones |
  4. | Stata 18: the mystery        1        18       7   Stata : the mystery |
     +------------------------------------------------------------------------+