type mismatch error on numeric variable

[Rachel Taylor]

**Editing to add: I resolved the problem! I don't know why but when I generated the problembinary variable it apparently was a string variable. Once I destringed that variable the commands worked. If anyone knows why it automatically was a string variable I am curious to know if I did something wrong there!


Hello, I am getting a type mismatch error on a numeric variable and can't figure out what I am doing wrong.

gen problembinary=.
replace problembinary=1 if c01_999==0
replace problembinary=0 if c01_999==1
replace problembinary=. if c01_999==.

The error comes after the first replace command. c01_999 is a numeric variable. If I describe it I get the following:
describe c01_999

storage display value
variable name type format label variable label
----------------------------------------------------------------------------------------------
c01_999 byte %8.0g 999 I have not experienced any problems


If I tab it I get the following:
tabulate c01_999

999 I have |
not |
experienced |
any |
problems | Freq. Percent Cum.
------------+-----------------------------------
0 | 1,355 33.49 33.49
1 | 2,691 66.51 100.00
------------+-----------------------------------
Total | 4,046 100.00


So I think the command should work, the only thing I can think of is something with the spacing? Anyone have any insights? I have Stata/SE 16.1 for Windows. Thanks!

[Andrew Musau]

Can you show us the exact commands and error message from Stata? Better still, copy and paste the output from

dataex in 1/20

[Nick Cox]

It's hard to say or see what went wrong here. For future reference

gen problembinary = 1 - c01_999

flips 0 and 1 and maps missings to missings.