st: RE: RE: RE: RE: Table of cut-offs for ROC following logit

["Martin Weiss" <martin.weiss1@gmx.de>]

<>

Glad it worked out for you! Does -rocfit- really have a -detail- option?


HTH
Martin


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Ploutz-Snyder, Robert (JSC-SK)[USRA]
Sent: Dienstag, 30. März 2010 18:59
To: statalist@hsphsun2.harvard.edu
Subject: st: RE: RE: RE: Table of cut-offs for ROC following logit

Thanks for the lead on Roger's senspec Martin.  That led me to roctab, which in combination with the predict post-estimation command to logit, works:

logit y var1 var2 var3 ...  vark			*to get the logit model

predict hat, pr  						*to get predicted probabilities of the outcome

rocfit y hat, detail					*generates the table of sens, spec, varying the cut-point 




-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Ploutz-Snyder, Robert (JSC-SK)[USRA]
Sent: Tuesday, March 30, 2010 11:44 AM
To: statalist@hsphsun2.harvard.edu
Subject: st: RE: RE: Table of cut-offs for ROC following logit

Doesn't that require a single predictor, plus dichotomous outcome?
I'm running a multiple logistic to predict the dichotomous outcome...



-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Martin Weiss
Sent: Tuesday, March 30, 2010 11:38 AM
To: statalist@hsphsun2.harvard.edu
Subject: st: RE: Table of cut-offs for ROC following logit


<>

" What am I missing??"



Roger`s

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

HTH
Martin

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Ploutz-Snyder, Robert (JSC-SK)[USRA]
Sent: Dienstag, 30. März 2010 18:29
To: statalist@hsphsun2.harvard.edu
Subject: st: Table of cut-offs for ROC following logit

I wish to obtain a table of cut-offs and associated sensitivity, specificity (and ppv, npv, fp, fn if possible) following a logit analysis.  I can use lsens to graph the relationship between sensitivity and specificity, given probability cutoffs shown on the x axis, but what I really want is the data that was used to create that lsens graph.

Seems this should be an option to lsens or perhaps estat class, but I can't locate it.

What am I missing??

Thanks much,
Rob

Robert J. Ploutz-Snyder, PhD 
Senior Research Scientist III (Biostatistician), University Space Research
Research Associate Professor, Department of Medicine, SUNY Upstate Medical University
Human  Adaptation & Countermeasures Division 
National Aeronautics and Space Administration 
Lyndon B. Johnson Space Center /SK 
2101 NASA Parkway 
Houston, TX 77058 
281-483-6296 (office) 
281-483-2888 (fax) 
Notice: This document, including any attachments, contains information that is confidential, protected by other applicable privileges, or constitutes non-public information. It is intended only for the designated recipient(s). If you are not an intended recipient of this information, please take appropriate steps to destroy this document in its entirety and notify the sender of its destruction. Use, dissemination, distribution, or reproduction of this information by unintended recipients is not authorized and may be unlawful. 
  Please consider the environment before printing this e-mail



*
*   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/