problem with looping the command tabstat

[Philipp Allroggen]

Hell everyone,

I want to observe the sd and mean of each of the countries in my panel dataset seperately to just get an overview.
Hence I used -tabstat- and ran a loop.

The dataset currently contains 11 countries, however it was originally larger (when I encoded it). Hence, the country variable does not run from 1 to 11, but from 1 to 23 with gaps.

To make it clearer, what I want to loop is:

tabstat $xlist if country1 == 1, stat(mean sd)
tabstat $xlist if country1 == 2, stat(mean sd)
tabstat $xlist if country1 == 8, stat(mean sd)
tabstat $xlist if country1 == 11, stat(mean sd)
.
.
.

I ran:

global xlist ca total_inv savings savinv nom_gdp_pc real_gdp_pc nom_catch_up catch_up nom_short_i real_short_i_pc real_short_i_gdp nom_long_i real_long_i_pc real_long_i_gdp tot reer_cp reer_tv erer mis_tv inflationpercentagechangeavrcp ncpi private_inv net_lending_borrowing lvl_dev depend depend_old depend_young unemply world_gdp_changes nom_us_gdp_pc fd_index

forvalue i == 1(1)23{
tabstat $xlist if country1 == `i', stat(mean sd)
}

However Stata stops after the first 2 countries or at the first gap, respectivly.
Interestingly, if I use -summarize- the code runs fine and just displays 0 as resluts when there is no country identifier.

forvalue i == 1(1)23{
summarize $xlist if country1 == `i'
}

But -summarize- gives unnecessary descriptives.

Any advise or hint is appreciated.

Thank you and stay healthy.

Philipp

[Nick Cox]

bysort country1 : tabstat $xlist, s(mean sd)

Here is some technique showing various ways to approach this.

. sysuse auto, clear
(1978 Automobile Data)

. bysort foreign : tabstat price mpg, s(mean sd)

--------------------------------------------------------------------------------------------------------------
-> foreign = Domestic

   stats |     price       mpg
---------+--------------------
    mean |  6072.423  19.82692
      sd |  3097.104  4.743297
------------------------------

--------------------------------------------------------------------------------------------------------------
-> foreign = Foreign

   stats |     price       mpg
---------+--------------------
    mean |  6384.682  24.77273
      sd |  2621.915  6.611187
------------------------------

foreach s in mean sd {
    foreach v in price mpg {
        egen `s'_`v' = `s'(`v') , by(foreign)
    }
}

tabdisp foreign, c(*_*)

----------------------------------------------------------------------
 Car type | Gear Ratio  mean_price    mean_mpg    sd_price      sd_mpg
----------+-----------------------------------------------------------
 Domestic |       3.58    6072.423    19.82692    3097.104    4.743297
  Foreign |       3.20    6384.682    24.77273    2621.915    6.611187
----------------------------------------------------------------------

. tabdisp foreign, c(*_price *_mpg)

----------------------------------------------------------
 Car type | mean_price    sd_price    mean_mpg      sd_mpg
----------+-----------------------------------------------
 Domestic |   6072.423    3097.104    19.82692    4.743297
  Foreign |   6384.682    2621.915    24.77273    6.611187
----------------------------------------------------------

. . egen tag = tag(foreign)

. . l foreign *_price *_mpg if tag, noobs

  +------------------------------------------------------+
  |  foreign   mean_p~e   sd_price   mean_mpg     sd_mpg |
  |------------------------------------------------------|
  | Domestic   6072.423   3097.104   19.82692   4.743297 |
  |  Foreign   6384.682   2621.915   24.77273   6.611187 |
  +------------------------------------------------------+

More technique at https://www.stata.com/support/faqs/d...-with-foreach/

[Philipp Allroggen]

Thank you very much for your answer!