Why the local value in a loop changes after use another data-set ?

[Eric Li]

I have two datasets. The first one contains the following data. The data name is Forloop.dta.

clear
input double(lpermno fyear) long apdedate float apdedateF90 str11 Station
10001 2010 18627 18537 "7252454853" 
10001 2011 18992 18902 "7252454853" 
10001 2012 19358 19268 "7252454853" 
10001 2013 19723 19633 "7252454853" 
10001 2014 20088 19998 "7252454853" 
10001 2015 20453 20363 "7252454853" 
10001 2016 20819 20729 "7252454853" 
10002 2003 16070 15980 "99819699999"
10002 2004 16436 16346 "99819699999"
10002 2005 16801 16711 "99819699999"
10002 2006 17166 17076 "99819699999"
10002 2007 17531 17441 "99819699999"
10002 2008 17897 17807 "99819699999"
10002 2009 18262 18172 "99819699999"
10012 2003 16130 16040 "72297793184"
end
format %d apdedate
format %td apdedateF90

The second dataset contains the following data. The dataset name is NOAP.dta .

clear
input str11 Station float(temp Date)
"72297793184" 56.7 17965
"72297793184" 67.9 21474
"72297793184" 59.1 16530
"72297793184" 63.6 17686
"72297793184" 59.7 18344
"72297793184" 62.6 16368
"72297793184" 57.7 18366
"72297793184" 58.8 17999
"72297793184" 61.4 19115
"72297793184" 61.4 21277
"72297793184" 71.1 18826
"72297793184" 61.1 19661
"72297793184" 69.4 21368
"72297793184" 62.4 20198
"72297793184" 65.9 16940
"72297793184" 63.9 19851
"72297793184" 66.4 14466
"72297793184" 60.6 17678
"72297793184" 64.8 17465
"72297793184" 60.3 14599
end
format %td Date

My code is :

clear
cd "C:\Data"
use "Forloop.dta"
local obs = _N
forvalue Num = 1(1)`obs'{
display `Num'
global StartDate=apdedateF90[`Num']
display $StartDate
global EndDate=apdedate[`Num']
display $EndDate
global StationName="Station[`Num']"
display $StationName
use "NOAP.dta"
display $StationName
keep if Station==$StationName
keep if inrange(Date,$StartDate,$EndDate)
cd "C:\TempDta"
save `Num',replace
cd "C:\Data"
use "Forloop.dta",clear
}

The example of the output is the following :

7252454853 *Before the use "NOAP.dta"
01001099999  *After the use "NOAP.dta"

My question is:
Why the value of the StationName is different before and after the use "NOAP.dta" ?
How to keep its value constant after use "NOAP.dta " ?
Thanks for your help.

[Joseph Coveney]

Why the value of the StationName is different before and after the[B] use "NOAP.[I]dta" ?

Putting the argument in quotation marks appears to assign the formula Station[`Num'] to the global macro instead of the value of the Numth station name in the dataset. So, when you display it each time, it evaluates the formula at that moment of display. The Numth value for station name differs between datasets; you get what the formula sees at the moment of evaluation.

How to keep its value constant after use "NOAP.dta " ?

Change the line to

global StationName = Station[`Num']

.ÿ
.ÿversionÿ15.1

.ÿ
.ÿclearÿ*

.ÿ
.ÿquietlyÿinputÿdouble(lpermnoÿfyear)ÿlongÿapdedateÿfloatÿapdedateF90ÿstr11ÿStation

.ÿformatÿ%tdCY-N-DÿapdedateÿapdedateF90

.ÿtempfileÿForloop

.ÿquietlyÿsaveÿ`Forloop'

.ÿ
.ÿdropÿ_all

.ÿquietlyÿinputÿstr11ÿStationÿfloat(tempÿDate)

.ÿformatÿ%tdCY-N-DÿDate

.ÿtempfileÿNOAP

.ÿquietlyÿsaveÿ`NOAP'

.ÿ
.ÿ/*ÿclear
>ÿcdÿ"C:\Data"
>ÿuseÿ"Forloop.dta"ÿ*/
.ÿuseÿ`Forloop'

.ÿlocalÿobsÿ=ÿ_N

.ÿforvaluesÿNumÿ=ÿ1(1)`obs'{
ÿÿ2.ÿÿÿÿÿÿÿÿÿdisplayÿ`Num'
ÿÿ3.ÿ
.ÿÿÿÿÿÿÿÿÿ/*ÿglobalÿStartDate=apdedateF90[`Num']
>ÿÿÿÿÿÿÿÿÿdisplayÿ$StartDate
>ÿ
>ÿÿÿÿÿÿÿÿÿglobalÿEndDate=apdedate[`Num']
>ÿÿÿÿÿÿÿÿÿdisplayÿ$EndDateÿ*/
.ÿ
.ÿÿÿÿÿÿÿÿÿglobalÿStationNameÿ=ÿStation[`Num']
ÿÿ4.ÿÿÿÿÿÿÿÿÿdisplayÿ"$StationName"
ÿÿ5.ÿ
.ÿÿÿÿÿÿÿÿÿ/*ÿuseÿ"NOAP.dta"ÿ*/
.ÿÿÿÿÿÿÿÿÿuseÿ`NOAP'
ÿÿ6.ÿÿÿÿÿÿÿÿÿdisplayÿ"$StationName"
ÿÿ7.ÿÿÿÿÿÿÿÿÿcontinue,ÿbreak
ÿÿ8.ÿ
.ÿÿÿÿÿÿÿÿÿ/*ÿkeepÿifÿStation==$StationName
>ÿÿÿÿÿÿÿÿÿkeepÿifÿinrange(Date,$StartDate,$EndDate)
>ÿÿÿÿÿÿÿÿÿcdÿ"C:\TempDta"
>ÿÿÿÿÿÿÿÿÿsaveÿ`Num',replace
>ÿÿÿÿÿÿÿÿÿcdÿ"C:\Data"
>ÿÿÿÿÿÿÿÿÿuseÿ"Forloop.dta",clearÿ*/
.ÿ}
1
7252454853
7252454853

.ÿ
.ÿexit

endÿofÿdo-file


.

[Eric Li]

Thanks Joseph. It really helps a lot.