Forvalues equal to the maximum

[Max Lin]

Hi Stata Users

I am using this code below and it is working very well:

bysort ID_com time: gen all=sum(v_local!=v_local[_n-1])
forval j=1/8{
egen v_local`j'=mean(cond(all==`j', v_local, .)), by (ID_com time)
}

and then I use this code which is also working very well:

gen byte trade_var= inlist(domestic,v_local1,v_local2,v_local3,v_local4,v_local5,v_local6,v_local7,v_local8)

my question is about

forval j=1/10

. In fact I do not know the exact number of "j"s. So I need to calculate first the maximum number of "j"s. The number could be 8, but also 10 or 12.
I tried using

forval j=1/max`j'

, which is obviously wrong and did not work. Is there any way to do that? And how would my second code change?

Do I have to provide data example for this question?

Thank you
Max

[Nick Cox]

For once, a data example isn't needed here.

su all, meanonly 

forval j = 1/`r(max)' {

should be enough of a hint, I hope.

[Max Lin]

Dear Nick Cox,

Thank you very much! It worked well.

I would like to know if it could also work with the other code. Assuming that the maximum of "all" was 3:
I have for example this code:

gen byte trade_var= inlist(domestic,v_local1,v_local2,v_local3)

and this code;

gen trade_var2 =1 if domestic==v_local & (v_local==v_local1 & v_local==v_local2) | (v_local==v_local1 & v_local==v_local3) |(v_local==v_local2 & v_local==v_local3)

they are both based on the code you have just helped me with.

Thank you
Max

[Nick Cox]

Sorry, I don't understand what you are seeking with #3.

[Max Lin]

After I run the code using `r(max)' as mentioned above, I will get: v_local1, v_local2, v_local3, v_local_n..... Since I do not have access to the data, I do not know the exact number of v_local. In this code I assume just as an example that they are 3.

gen trade_var=1 if domestic == v_local1 | domestic == v_local2 | domestic == v_local3
*or as I wrote above 
 gen byte trade_var= inlist(domestic,v_local1,v_local2,v_local3)

My question is how I could also use `r(max)' in this code. and the other code mentioned in #3.

Thanks

[Nick Cox]

Fair enough. You can store r(max) any way you wish

local max = r(max)

scalar max = r(max)

gen max = r(max)

with varying longevity.

You can also keep track of the variables created adding a line within the loop

local newvars `newvars'  v_local`j'

Then outside the loop you have newvars containing

v_local1 v_local2 v_local3

or whatever. If you need those names to be comma-separated, that is as simple as

local newvars_c : subinstr local newvars " " ",", all

and benefit by

 
 gen byte trade_var = inlist(domestic, `newvars_c')

[Max Lin]

Dear Nick,

Thank you very much for your help!

Best wishes
Max

[Max Lin]

Sorry, I just had an issue with

 
 gen byte trade_var = inlist(domestic, `newvars_c')

 
 gen byte trade_var = inlist(domestic, `newvars_c') invalid syntax
r(198);

end of do-file

Data example is as follows:

* Example generated by -dataex-. For more info, type help dataex
clear
input int(year id_com v_local domestic)
2005 19587 111 111
2005 19587 111 112
2005 19587 111 113
2005 19587 111 114
2005 19587 112 111
2005 19587 112 112
2005 19587 112 113
2005 19587 112 114
2005 19587 114 111
2005 19587 114 112
2005 19587 114 113
2005 19587 114 114
2005 19587 114 111
2005 19587 114 112
2005 19587 114 113
2005 19587 114 114
2005 19587 111 111
2005 19587 111 112
2005 19587 111 113
2005 19587 111 114
2005 19587 111 111
2005 19587 111 112
2005 19587 111 113
2005 19587 111 114
end

[Nick Cox]

You need to put something like

macro list

or specifically

di "`newvars_c'"

before the command that fails to see what is being objected to.

One possibility is that you're forgetting that local macros have only local scope: see

SJ-20-2 dm0102 . . . . . . . . . Stata tip 138: Local macros have local scope
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . N. J. Cox
Q2/20 SJ 20(2):499--503 (no commands)
focuses on a common misunderstanding of how local macros work

if you have access.

[Max Lin]

The code works well! Thanks.
However, I guess the code is correct but what I need to generate was something different. Th three codes below, as I understood, should give the same results:

 
 *Code 1: gen trade_var=1 if domestic == v_local1 | domestic == v_local2 | domestic == v_local3 replace trade_var=0 if trade_var==. *Code 2:  gen byte trade_var= inlist(domestic,v_local1,v_local2,v_local3)

give:

* Example generated by -dataex-. For more info, type help dataex
clear
input int year byte(id_com v_local domestic all v_local1 v_local2 v_local3 v_local4 trade_var)
2005 1 11 11 1 11 12 14 11 1
2005 1 11 12 1 11 12 14 11 1
2005 1 11 13 1 11 12 14 11 .
2005 1 11 14 1 11 12 14 11 1
2005 1 12 11 2 11 12 14 11 1
2005 1 12 12 2 11 12 14 11 1
2005 1 12 13 2 11 12 14 11 .
2005 1 12 14 2 11 12 14 11 1
2005 1 14 11 3 11 12 14 11 1
2005 1 14 12 3 11 12 14 11 1
2005 1 14 13 3 11 12 14 11 .
2005 1 14 14 3 11 12 14 11 1
2005 1 14 11 3 11 12 14 11 1
2005 1 14 12 3 11 12 14 11 1
2005 1 14 13 3 11 12 14 11 .
2005 1 14 14 3 11 12 14 11 1
2005 1 11 11 4 11 12 14 11 1
2005 1 11 12 4 11 12 14 11 1
2005 1 11 13 4 11 12 14 11 .
2005 1 11 14 4 11 12 14 11 1
end

 
 *Code 3:  gen byte trade_var = inlist(domestic, `newvars_c')

gives

* Example generated by -dataex-. For more info, type help dataex
clear
input int year byte(id_com v_local domestic all v_local1 v_local2 v_local3 v_local4 trade_var)
2005 1 11 11 1 11 12 14 11 1
2005 1 11 12 1 11 12 14 11 0
2005 1 11 13 1 11 12 14 11 0
2005 1 11 14 1 11 12 14 11 0
2005 1 12 11 2 11 12 14 11 0
2005 1 12 12 2 11 12 14 11 1
2005 1 12 13 2 11 12 14 11 0
2005 1 12 14 2 11 12 14 11 0
2005 1 14 11 3 11 12 14 11 0
2005 1 14 12 3 11 12 14 11 0
2005 1 14 13 3 11 12 14 11 0
2005 1 14 14 3 11 12 14 11 1
2005 1 14 11 3 11 12 14 11 0
2005 1 14 12 3 11 12 14 11 0
2005 1 14 13 3 11 12 14 11 0
2005 1 14 14 3 11 12 14 11 1
2005 1 11 11 4 11 12 14 11 1
2005 1 11 12 4 11 12 14 11 0
2005 1 11 13 4 11 12 14 11 0
2005 1 11 14 4 11 12 14 11 0
end

[Nick Cox]

I can't really comment on what else you might need.