How to split string variables some separated by commas and without commas or space

[Chris Miyinzi]

Dear all,

I would like to split string variables for instance if the original variable has an observation such as "p2.1, p3.1 and p4.1", I would like to split it in such a way that three variables are generated
with values (2.1, 3.1 and 4.1) and should be numeric.

Please find sample dataset that I am working on below

* Example generated by -dataex-. To install: ssc install dataex
clear
input str20(m10_10_4_livefences m10_10_4_terracing m10_10_4_mintillage m10_10_4_farmmanure)
""     "P2.1,P3.1,P4.1" ""     ""
""     "P2.1"           ""     ""
""     "P2.1"           ""     ""
""     "P2.1,P3.1"      ""     ""
""     "P2.1,P3.1"      ""     ""
""     "P1.2,P2.1"      "P3.1" ""
""     "P2.1,P3.1"      "P4.1" ""
""     "P3.1,P4.1"      ""     ""
""     "P2.1,P3.1,P4.1" ""     ""
""     ""               ""     ""
"P1.1" ""               ""     ""
end

Thanks in Advance

[Jorrit Gosens]

Short answer:

split m10_10_4_terracing, p(",")

However, this seems as if the data has imply not been imported correctly. If that's the case, perhaps showing a snippet of the original data and the command you used to import is useful.
Further: if you do the code above, it will create a set of new variables; Stata cannot decide for you if some of these new values should actually have come as values under any of the existing variables (or which)

[Chris Miyinzi]

Thanks a lot Jorrit. The data was imported from CS-Pro. After Spllting and I get a variable with an element like p2.1 is it possible to split it further i generate two variables one with p and another one with 2.1. Is it posible?

[Jorrit Gosens]

What variations are there? Is it always only a single letter before the numeric part?

If it is always just one letter, follwoing the split as in the example above, do:

gen letter1 = substr( m10_10_4_terracing1, 1, 1)
gen number1 = subinstr( m10_10_4_terracing1 , letter1, "",.)

The first line creates a variable containing the first character of the original string variable, the second line removes the character defined in the first line from your original string variable

[Chris Miyinzi]

Thanks a lot @ Jorrit It worked perfectly.

[Gupta Manish]

Thank you, Jorrit. It worked for me as well after more than 3 years.

Manish