Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: xi generated varnames

From   Lars Korsholm <>
Subject   st: xi generated varnames
Date   Fri, 15 Nov 2002 15:20:53 +0100 (MET)


I want to get a hold on a xi generated variable
however depending on the length of the variable the structure of the xi
generated name differs. Any help

here follows my details:
the program take just one argument `stub' which indicate which set of
variables to run with

xi: regress `stub'_fu  dosis_F dosis_M i.dosis_gruppe|`stub'_bl,  robust
     di "* test of different baseline effekt in three treatment groups *"
     qui test  _IdosX`stub'_b_2= _IdosX`stub'_b_3, notest
     test  _IdosX`stub'_b_2= _IdosX`stub'_b_4, accum

I first worte and tejcked this program when `stub' was a three letter
combination and the the _Idos... worked fine
now I have a stub wiht five letters (extfat) which by xi gives variables
on the form  _IdosXextfa_2 etc
(note that the last "t" in extfat is choped off so even _IdosX`stub'_2
dont exist)

My real problem is that I have 4 dosis group. I want to test wheter the
`stub'_bl effect is the same in group 2,3,4 (excluding 1 (placebo))

Hope someone might help

Best lars

PS I have tried to 
xi: regress .... if dosis_group!=1 ...
but then `stub'_bl is excluded by stata
test _IdosX*
dont test the right hypothesis. Its test whether all effects are 0, not

: Lars Korsholm Ph.D.,                     :
:  Department of Statistics and Demography :
:  University of Southern Denmark          :
:  Sdr. Boulevard 23A, 1.                  :
:  DK-5000 Odense C                        :
: Phone   +45 6550 3608                    :
: Fax     +45 6595 7766                    :
: Mail              :
: Web  :

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index