Usijng split

[Euslaner]

I have a date 4/25/2000 and need to divide it to three variables 4 25 and 2000. Can I do this with split and if so how.

tyoed. edit var11

. edit var1

. split var1,gen(day month year)
invalid stub day month year

and received:

r(198);


Any help would be appreciated.

Ric

[Clyde Schechter]

split var1, parse("/") gen(dmy) destring
rename (dmy*) (day month year)
drop dmy*

[Euslaner]

Thanks Nick

[Nick Cox]

I should perhaps do something to deserve that after Clyde Schechter solved the problem. Perhaps the credit arises from my original contribution on split.

Ric (Eric) has a problem with a daily date in string form. So we could call up date functions too, as in

* Example generated by -dataex-. For more info, type help dataex
clear
input str9 ddate
"4/25/2000"
"7/1/2025" 
end

foreach f in month day year {
    gen `f' = `f'(daily(ddate, "MDY"))
}

list 

    +--------------------------------+
     |     ddate   month   day   year |
     |--------------------------------|
  1. | 4/25/2000       4    25   2000 |
  2. |  7/1/2025       7     1   2025 |
     +--------------------------------+

For an introduction to Statalist, including dataex, please read https://www.statalist.org/forums/help

For a more recent tutorial survey of dates, please read https://journals.sagepub.com/doi/pdf...6867X251341416

gen ndate = daily(ddate, "MDY") 
format ndate %td

could also be useful.

[Euslaner]

Sorry Clyde I meant to thank you,. But I have another problem now. I have data on GDP per capita over time and a line of data reads:

1871 | 2489 | 1.79%

I want to divide the data into separate variables with year first and GDP last with the % sign off. I typed, following your suggestion as I read it:

. split var1, parse("/") gen(yxg)
cannot generate new variables using stub yxg
r(110);

What did I do wrong? Any help would be appreciated.

[Nick Cox]

You wrote forward slash / but meant to write pipe |

http://www.catb.org/jargon/html/A/ASCII.html lists various names for characters.

[Chen Samulsion]

You must have a variable with stub yxg in your data. And pipe | sign is certainly a problem too.

. set obs 1
number of observations (_N) was 0, now 1

. generate str var1 = "1871 | 2489 | 1.79%"

. split var1, parse("/") gen(yxg)
variable created as string: 
yxg1

. split var1, parse("/") gen(yxg)
cannot generate new variables using stub yxg
r(110);