Thanks Martin - you've got me moving forward again.
You wouldn't happen to know how to get the model pseudo r-square
displayed would you? In Roger's post below he suggests that parmby
calculates and stores it as r2_p, but I don't see it.
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Martin Weiss
Sent: Saturday, June 27, 2009 3:44 PM
To: [email protected]
Subject: st: Re: RE: RE: R2 stats using statsby or parmby???
<>
Not sure whether that blank after -flist-- hurt you, but you should use
line
continuation "///". Also note that -by- likes to be passed a dataset
sorted
by its -varlist-, so append an "s" to make it "bysort"...
***
use http://www.ats.ucla.edu/stat/stata/dae/logit.dta, clear
global tflist ""
global modseq=0
foreach X in gre topnotch gpa {
global modseq=$modseq+1
tempfile tf$modseq
parmby "logit admit `X'", label ///
command format(estimate min95 max95 ///
%8.2f p %8.1e) idn($modseq) ///
saving(`tf$modseq',replace) ///
flist(tflist)
}
dsconcat $tflist
sort idnum parmseq
describe
bys idnum command:list parm ///
label estimate min95 max95 p,noobs
***
HTH
Martin
_______________________
----- Original Message -----
From: "Hugh Robinson" <[email protected]>
To: <[email protected]>
Sent: Saturday, June 27, 2009 11:10 PM
Subject: st: RE: RE: R2 stats using statsby or parmby???
Roger,
I was intrigued by reading your response to this posting, I think the
foreach and parmby commands could be very helpful for me in running a
number of univariat logits and producing a single table containing
each
variable name, its coefficient, p-value, and r-square value.
Towards that goal I thought I would start with the UCLA sample logit
data and try to modify the command list from the parmby help file to
create something looks like what I'm after.
When I run the following code the foreach loop only gets as far the
first covariate "gre" then produces an error that reads "variable
tflist
not found".
Can you see what's wrong?
Thanks
HR
use http://www.ats.ucla.edu/stat/stata/dae/logit.dta, clear
global tflist ""
global modseq=0
foreach X in gre topnotch gpa {
global modseq=$modseq+1
tempfile tf$modseq
parmby "logit admit `X'", label command format(estimate min95 max95
%8.2f p %8.1e) idn($modseq) saving(`tf$modseq',replace)
flist (tflist)
}
dsconcat $tflist
sort idnum parmseq
describe
by idnum command:list parm label estimate min95 max95 p,noobs
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Newson,
Roger
B
Sent: Friday, June 26, 2009 1:07 PM
To: '[email protected]'
Subject: st: RE: R2 stats using statsby or parmby???
You do not specify what goodness-of-fit statistic you want, or which
program you are using to do your logistic regression (O prefer -glm-,
but many other people prefer -logit- or -logistic-). However, if the
goodness-of-fit statistic is saved in e(), then either -statsby- or
-parmby- can save it. In the case of -parmby-, you use the -escal()-
option to save extra scalar results. As in
parmby "glm y x, link(logit) family(bin) eform robust", eform
norestore
by(group) escal(aic bic chi2 dispers_p)
which should save -e(aic)-, -e(bic)-, -e(chi2)- and -e(dispers_p)- in
scalar variables named -es_1-, -es_2-, -es_3-, and -es_4-,
respectively.
You can give them more informative names using the -rename()- option.
Similarly, the -logit- and -logistic- commands save the
pseudo-R-squared
in -e(r2_p)-.
I hope this helps.
Best wishes
Roger
Roger B Newson BSc MSc DPhil
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton Campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
UNITED KINGDOM
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Email: [email protected]
Web page: http://www.imperial.ac.uk/nhli/r.newson/
Departmental Web page:
http://www1.imperial.ac.uk/medicine/about/divisions/nhli/respiration/pop
genetics/reph/
Opinions expressed are those of the author, not of the institution.
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Thomas M
Holbrook
Sent: 26 June 2009 18:09
To: [email protected]
Subject: st: R2 stats using statsby or parmby???
I'm trying to find a way the generate goodness-of-fit stats for logit
models run separately over subgroups of my my data (I'm running a vote
model using individual-level data and I want to generate fit stats by
day of the campaign). I can get the slopes and standard errors using
"statsby" or "parmby" but I don't set a way of generating the fit
stats.
Any ideas???
-Tom
Thomas M. Holbrook
Wilder Crane Professor of Government
Department of Political Science
University of Wisconsin-Milwaukee
3210 North Maryland Avenue
Milwaukee, WI 53211
www.uwm.edu/~holbroot
www.election08data.blogspot.com
414-229-6468
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
