Data Setting: dropping empty variables in a quickly way

[Chiara Tasselli]

Good morning to everyone,
i was wondering if there is a command that tells Stata to drop variables if contains only missing values.

I have a database with numerous variables and I would like to delete the ones that are completely empty without having to check them one by one.

Many thanks in advance for your time.
Wishing you a happy weekend.

[Rich Goldstein]

you want the user-written -missings- command (use -search- of -findit- to locate and install); when looking at the help file you will see that one of the sub-commands is "dropvars" and that is what you want

[Andrew Musau]

With the -missing()- function, this is easy as you do not need to differentiate between numeric variables and strings.

foreach var of varlist *{
    sort `var'
    if missing(`var'[1]) & missing(`var'[_N]){
        drop `var'
    }
}

[Nick Cox]

With findname (Stata Journal) you can go

findname, all(missing(@)) 
drop `r(varlist)'

[Chiara Tasselli]

thank you all very much,
I received lots of useful advice and solved my problem.

[li zhi]

With findname (Stata Journal) you can go

findname, all(missing(@))
drop `r(varlist)'

hi,Nick! Can you explain the meaning Of @ for me? I don't understand this syntax.

[Nick Cox]

See the help. Any condition used in an all() or any() option is a true-or-false condition defined by an expression in which variable
names are represented by @. For example, any(@ < 0) selects numeric variables in which any values are
negative.

[li zhi]

See the help. Any condition used in an all() or any() option is a true-or-false condition defined by an expression in which variable
names are represented by @. For example, any(@ < 0) selects numeric variables in which any values are
negative.

Is it only applicable in this command？I mean ，does it relate to matrix or something else,for example, it represents all the values of one variable?

[Nick Cox]

@ syntax is not generally available. If it is, that will be documented in the help. Nor does @ syntax always mean the same thing. Compare mylabels from SSC and coefplot from Stata Journal and SSC.

[li zhi]

@ syntax is not generally available. If it is, that will be documented in the help. Nor does @ syntax always mean the same thing. Compare mylabels from SSC and coefplot from Stata Journal and SSC.

I see, thank you very much!