strofreal vs sprintf

[Sergiy Radyakin]

In Stata (and Mata) there are left-justified and right-justified formats for numbers. For some reason strofreal() function does not obey the change in justification while the sprintf() works as I expected (Stata 13.0). I wonder if this is a bug or is there a reason for this. Compare for illustration:

mata "<"+strofreal(7,"%-2.0f")+">"
 mata "<"+strofreal(7,"%2.0f")+">"
 mata sprintf("<%2.0f>",7)
 mata sprintf("<%-2.0f>",7)

(the angle brackets are added for clarity)

I am interested in which format will deliver the behavior with leading spaces (not zeroes)? and whether the behavior of strofreal() or sprintf() is more intuitive?

Thank you,
Sergiy Radyakin

[Joe Canner]

Sergiy,

According to the strofreal() documentation:

Leading blanks are trimmed from the result.

So, it's not a bug, although it's not clear what the reason is. My guess this is a philosophical choice based on the assumption (probably generally correct) that users would want a string with just a number, with no blanks. If anything, the fact that your first example still has trailing blanks might be considered a bug, in that such a philosophy would dictate that trailing blanks should also be trimmed, but perhaps the use of %-2.0f was not anticipated.

Regards,
Joe

[Sergiy Radyakin]

Dear Joe, thank you for your answer. I looked very thoroughly throw the documentation for the formats, but couldn't find sufficient explanation. With your reference I see that no matter what I do with formats, it is the strofreal() function that distorts the result and it is by design (still no idea why). As I do need leading blanks, I will have to use sprintf() as a workaround. Thank you, Sergiy.