Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: RE: RE: RE: RE: RE: change ROC Curve default


From   "Martin Weiss" <[email protected]>
To   <[email protected]>
Subject   st: RE: RE: RE: RE: RE: change ROC Curve default
Date   Tue, 16 Mar 2010 09:25:28 +0100

<>

*************
ssc d senspec
*************

by Roger Newson may also be a good tool for you while experimenting with
this issue...



HTH
Martin

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Sheri Stevens
Sent: Dienstag, 16. März 2010 01:58
To: [email protected]
Subject: st: RE: RE: RE: RE: change ROC Curve default

Thank you very much!
Sheri

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Martin Weiss
Sent: Monday, March 15, 2010 8:16 PM
To: [email protected]
Subject: st: RE: RE: RE: change ROC Curve default


<>

-lroc- depicts all kinds of cutoffs. So it is hard to change anything here,
but you can highlight with -xline- and -yline-:


*************
sysuse auto, clear
logit foreign length weight
loc i 0
qui forv j=.1(.1).9{
loc ++i
estat clas, cut(`j')
lroc, xli(`=(100-r(P_n0))/100') yli(`=r(P_p1)/100') name(gr`i', replace)
nodraw
}
graph combine gr1 gr2 gr3 gr4 gr5 gr6 gr7 gr8 gr9
*************


If ever you wanted to see the entire basis for the graph, -roctab- is your
friend:

*************
logit foreign length weight
predict pred
roctab foreign pred, detail
*************

HTH
Martin


-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Sheri Stevens
Sent: Dienstag, 16. März 2010 00:47
To: [email protected]
Subject: st: RE: RE: change ROC Curve default

Dear Martin, thank you, the classification table information is very
helpful!!  I was hoping to change the default cut point of the ROC Curve,
which I was told is .5. I did check out the help menu but couldn't find an
option that would do that.
Sheri

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Martin Weiss
Sent: Monday, March 15, 2010 7:20 PM
To: [email protected]
Subject: st: RE: change ROC Curve default


<>

Which "default .5" are you referring to? The help file -help
logistic_postestimation- shows you all available options. Maybe you want the
classification table where you can indeed change the cutoff?



*************
sysuse auto, clear
logit foreign length weight
estat clas, cut(.8)
estat clas, cut(.2)
*************


HTH
Martin

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Sheri Stevens
Sent: Montag, 15. März 2010 22:51
To: [email protected]
Subject: st: change ROC Curve default

When generating a ROC Curve in logistic regression post estimation I would
like to be able to change the default of .5 and was wondering if there is an
option in Stata to do that.  Any help would be most appreciated!
The command I am using is: 
lroc

Thank you,

Sharon Stevens, LMSW
Community and Public Service Program
Social Sciences 112
SUNY Albany
1400 Washington Ave.
Albany, NY 12222
518.442.5683     fax 518.442.5684



*
*   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/


*
*   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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index