String variable to Numeric variable

[Marco Errico]

Dear everyone,

I'm working on a large dataset that comes from a regulator. Some variables are csv and I imported in stata.
I have two main issues and one is the following. I have a string variable that I would like to convert in numeric one where "," is decimal separator and "." is thousand separator.
I'm wondering what could be the best approach to convert in a simple stata numeric variable

I tried the

real

function but it simplies give me missing values for all of them.
I also explored the properties of the variable:
Type: str9
Format: %9s

Someone that would be so kind with any advice?
Thank you in advance

* Example generated by -dataex-. To install: ssc install dataex
clear
input str9 adnt2017
"0,40"    
"30,73"   
"0,09"    
"66,82"   
"71,04"   
"0,04"    
"0,01"    
"0,03"    
"0,28"    
"134,67"  
"29,99"   
"197,51"  
"0,26"    
"207,58"  
"265,90"  
"2.321,99"
"502,89"  
"2,14"    
"1,19"    
"2,22"    
"0,55"    
"13,51"   
"2,71"    
"0,33"    
"0,28"    
"0,74"    
"743,14"  
"1.287,56"
"3.107,41"
"266,04"  
"573,07"  
"0,22"    
"112,46"  
"0,09"    
"532,67"  
"0,40"    
"0,25"    
"0,57"    
"1.497,67"
"0,00"    
"194,11"  
"1,21"    
"2.754,51"
"696,78"  
"0,30"    
"0,44"    
"179,96"  
"1,93"    
"1,21"    
"0,71"    
"4,80"    
"17,76"   
"198,40"  
"0,32"    
"0,98"    
"0,89"    
"1.389,80"
"41,11"   
"0,48"    
"464,54"  
"28,22"   
"1.050,54"
"0,02"    
"38,68"   
"1.411,40"
"4,46"    
"613,49"  
"0,47"    
"767,94"  
"36,02"   
"2.278,11"
"154,79"  
"710,48"  
"246,48"  
"194,41"  
"38,89"   
"0,21"    
"1,07"    
"866,90"  
"0,28"    
"0,03"    
"0,50"    
"0,71"    
"0,00"    
"1.143,77"
"2,34"    
"756,73"  
"956,82"  
"102,52"  
"1.433,31"
"0,00"    
"0,02"    
"0,00"    
"0,00"    
"322,59"  
"0,07"    
"0,18"    
"0,00"    
"0,00"    
"0,00"    
end

[Øyvind Snilsberg]

remove thousand separator,

replace adnt2017 = subinstr(adnt2017,".","",.)

replace decimal separator,

replace adnt2017 = subinstr(adnt2017,",",".",.)

destring,

destring(adnt2017), replace

[Rich Goldstein]

the following works on your example data:

destring adnt2017, gen(adnt2017b) dpcomma ignore(".")

see

help destring

[Maarten Buis]

// this does the work
destring adnt2017, gen(adnt2017_num) dpcomma ignore(".")

// cosmetic, just add a 1000s mark to make it easier to read
format adnt2017_num %9.2fc

// admire the result
list

[Marco Errico]

Thanks Øyvind Snilsberg and Rich Goldstein

I finally got a solution with which is exactly what you rich suggest

I did think about to play with destring function.

Thanks again for your time!