How to find the numeric values that stand for different meanings of string variables?

[Zhang Hui]

I have met a problem. I have a dataset. There is a string variable named country. I firstly destringed it and created a new variable labeled country_num.

Now I want to find the corresponding numeric values of some countries, like US, UK and China and so on. For example, if I want to search some numbers of countries, I have to find their string names and corresponding numbers one by one.

I tired it by using the command,

summarize country_num if country == "United States"|"United Kingdom"

.

But it failed.

How can I search it quickly? Would you like to help me? Thanks a lot.

[Carlo Lazzaro]

Zhang:

label list country_num

is the first step to take.

[Zhang Hui]

Dear Carlo, thanks for your help!

[Zhang Hui]

Thanks again.

[Carlo Lazzaro]

Zhang:
you're welcome.

[Nick Cox]

Some confusion here.

First, there is an allusion to destring, but destring would not turn a string variable with values like "Albania" into a numeric variable with those values as value labels. That is what encode does.

Second, the problem with

summarize country_num if country == "United States"|"United Kingdom"

is quite different. Stata parses the argument to if as

* fine! 
country == "United States"

* fine! 
|

* not valid in this context 
"United Kingdom"

and the problem here is that the last is not a true-or-false expression (political jokes aside, and no, I didn't vote stupidly). What would be legal is

summarize country_num if country == "United States"| country == "United Kingdom"

or

summarize country_num if inlist(country, "United States", "United Kingdom")

Naturally, doing it with numeric values for the encoded variable is fine too, but not the only way to proceed.

[Zhang Hui]

Thanks for your help!