Destringing a variable

[Tariq Abdullah]

I'm trying to destring my totaltrade variable but it's returning me the following error.

My dataset looks like this :

* Example generated by -dataex-. For more info, type help dataex
clear
input str12 totaltrade
"284,944,468"
"237,956,928"
"136,575,034"
"111,223,976"
"101,089,801"
"92,208,560" 
"73,400,104" 
"63,004,373" 
"61,701,612" 
"58,038,313" 
"56,893,814" 
"6,522,640"  
"8,838,236"  
end

My code is this

 destring  totaltrade, replace ignore(" ")

Getting this error after putting that command

totaltrade: contains characters not specified in ignore(); no replace

Any helpful suggestion is going to be very much appreciated

[Joro Kolev]

Most probably you mean

 
 destring  totaltrade, replace ignore(",")

[Tariq Abdullah]

Thanks so much for the kind suggestion, But, still the issue prevails.

destring  totaltrade, replace ignore(",")
totaltrade: contains characters not specified in ignore(); no replace

[Nick Cox]

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

tab totaltrade if missing(real(totaltrade))

may show you further non-numeric characters, such as "NA" for missing or metadata in the data.

[Tariq Abdullah]

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

tab totaltrade if missing(real(totaltrade))

may show you further non-numeric characters, such as "NA" for missing or metadata in the data.

After applying it, this got rid of the problem of commas. But, I had the intention of converting the variable into integer. The last time, I had a similar variable as string , and after destringing it, it was turned into numerous one after applying my #1 post command.

However, this time despite your command, I can see it's still a string variable. I apologize for not being clear about making it numeric from string (because I thought by destringing it I'll be able to turn it into integer like last time I had to experience a similar type of variable ).

 
"6039927"  
"5500928"  
"14837762" 
"12205883" 
"4763728"  
"5248146"  
"6522640"  
"8838236"  
end

Would you be kind enough one more time to suggest me how I can transform this string variable into integer ?

[Nick Cox]

You are not showing the results of

 tab totaltrade if missing(real(totaltrade))

which is intended to show -- apart from commas -- what is the problem. Then you can apply destring. with appropriate options.

[Joro Kolev]

You are not trying to understand your problem, but instead you are trying to apply something that will automatically resolve your problem.

The command you need in principle is

 
 destring  totaltrade, replace ignore(",")

It works on the sample you showed as it should. If it does not work on your full data, it means that you have other symbols there, and you need to figure out what are these symbols, and why do you have them.

Type for example in the lines previously suggested

tab totaltrade if missing(real(totaltrade))

and this will show you the problematic values of your variable. a

[Tariq Abdullah]

You are not showing the results of

tab totaltrade if missing(real(totaltrade))

which is intended to show -- apart from commas -- what is the problem. Then you can apply destring. with appropriate options.

tab totaltrade if missing(real(totaltrade))

 total trade |      Freq.     Percent        Cum.
-------------+-----------------------------------
          n0 |          1      100.00      100.00
-------------+-----------------------------------
       Total |          1      100.00

This is what it shows and my apologies for missing out on that

[Joro Kolev]

If I were you, I would do the following:

 
 destring  totaltrade, gen(totaltradenumeric) force ignore(",")

some of your data will now go to missing, the data with the mysterious symbols inside. To inspect this data, and decide what to do you can do

tab totaltrade if missing(totaltradenumeric)

The next actions will depend on what you see from the last tabulation.

[Tariq Abdullah]

As soon as I fixed the value with n0 The command you have kind suggested worked out fine. Thanks so much for being so patient with me throughout the whole time ! Much appreciated