Problem with lag and lead operators in xtset

[Gianni De Fraja]

I have a problem with the F. command in xtset. The relevant part of the dataset is:

. xtset authid year
. d authid year afid countrycode

Variable      Storage   Display    Value
    name         type    format    label      Variable label
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
authid          double  %15.0g                Scopus identifier of author
year            float   %9.0g                
afid            long    %15.0g                Scopus identifier of affiliation.
countrycode     int     %52.0g     etich      country-code

For a given panel, I get:

. list year  authid afid* countrycode* if authid == 7003989689

         +-----------------------------------------------+
         | year       authid       afid      countrycode |
         |-----------------------------------------------|
 549793. | 1988   7003989689   60016418   United Kingdom |
 549794. | 1989   7003989689   60033125   United Kingdom |
 549795. | 1990   7003989689   60033125   United Kingdom |
 549796. | 1991   7003989689   60033125   United Kingdom |
 549797. | 1992   7003989689   60033125   United Kingdom |
         |-----------------------------------------------|
 549798. | 1993   7003989689   60033125   United Kingdom |
 549799. | 1994   7003989689   60033125   United Kingdom |
 549800. | 1995   7003989689   60015138   United Kingdom |
 549801. | 1996   7003989689   60015138   United Kingdom |
 549802. | 1997   7003989689   60015138   United Kingdom |
         |-----------------------------------------------|
 549803. | 1998   7003989689   60015138   United Kingdom |
 549804. | 1999   7003989689   60027509            Italy |
         +-----------------------------------------------+

I want to generate a variable containing the next year's countrycode and authid. The commands

. gen countrycode_next = F.countrycode
(2,901,342 missing values generated)

. gen afid_next = F.afid
(2,901,342 missing values generated)

give the following outcome:

. list year  authid afid* countrycode* if authid == 7003989689
 
         +-------------------------------------------------------------------------------------------------------------------------------+
         | year       authid       afid   afid_n~t   afid_s~o      countrycode   countrycode_~t |
         |--------------------------------------------------------------------------------------------------------------------------|
 549793. | 1988   7003989689   60016418   60033124          .   United Kingdom   United Kingdom |
 549794. | 1989   7003989689   60033125   60033124          .   United Kingdom   United Kingdom |
 549795. | 1990   7003989689   60033125   60033124          .   United Kingdom   United Kingdom |
 549796. | 1991   7003989689   60033125   60033124          .   United Kingdom   United Kingdom |
 549797. | 1992   7003989689   60033125   60033124          .   United Kingdom   United Kingdom |
         |---------------------------------------------------------------------------------------------------------------------|
 549798. | 1993   7003989689   60033125   60033124          .   United Kingdom   United Kingdom |
 549799. | 1994   7003989689   60033125   60015136          .   United Kingdom   United Kingdom |
 549800. | 1995   7003989689   60015138   60015136          .   United Kingdom   United Kingdom |
 549801. | 1996   7003989689   60015138   60015136          .   United Kingdom   United Kingdom |
 549802. | 1997   7003989689   60015138   60015136          .   United Kingdom   United Kingdom |
         |----------------------------------------------------------------------------------------------------------------|
 549803. | 1998   7003989689   60015138   60027508          .   United Kingdom            Italy |
 549804. | 1999   7003989689   60027509          .          .            Italy                . |
         +-----------------------------------------------------------------------------------------------------+

While the countrycode is correct, the afid is reduced by 1 or 2. I cannot think of an error which would determine this behaviour, and any help would be much appreciated (PS this is my first post, so apologies if I have inadvertently missed some convention).

[Andrew Musau]

See

help precision

The default storage type for numeric variables in Stata is float, and your numbers are too big for this. See, e.g.,

di float(60016418)

Res.:

. di float(60016418)
60016416

So you want to hold the generated variable in double precision.

gen double afid_next = F.afid

[William Lisowski]

Also, looking at post #1, we see that afid was stored as a long variable, which should have served as a hint that F.afid should at least be stored as a long variable. But as the table below tells us, a double has more precision than a long, so nothing is lost by storing it in a long.

Here are the limits on storage of decimal integers with full accuracy in the various numeric storage types. The fixed-point variables lose the 27 largest positive values to missing value codes; the similar loss for floating point variables occurs only for the largest exponent, so it doesn't affect the much smaller integer values.

byte - 7 bits	-127	100
int - 15 bits	-32,767	32,740
long - 31 bits	-2,147,483,647	2,147,483,620
float - 24 bits	-16,777,216	16,777,216
double - 53 bits	-9,007,199,254,740,992	9,007,199,254,740,992

[Gianni De Fraja]

Thank you both very much. It works a treat, and I can see the reason for the error.