[R] Clogit R and Stata

2013-06-07 Thread Richard Beckett 
Sorry to once again write a message but I'm once again stumped and am having no 
luck finding a solution anywhere else.


This question requires some finesse in both R and STATA so hopefully I will be 
able to get an answer here. I am much more adept in R and am trying to 
replicate the results of a STATA file in R. Hopefully this is a proper forum 
for such questions. 


This is the code for the clogit in STATA
clogit sftpcons sftptv2a3 sftptv2a4 sftptv2a5 sftptv2a2 sftptv2a6 logim maccat 
disp4cat if sample==1  glb_ind==Y, group(stratida)
and I tried to replicate it using
clogit1-clogit(sftpcons~sftptv2a3+sftptv2a4+sftptv2a5+sftptv2a2+sftptv2a6+logim+maccat+disp4cat+strata(stratida),
 dframe, sample==1 | glb_ind==Y)
but got different results
What did I do wrong here? I interpreted the STATA clogit as run this logit as 
long as the sample is 1 and glb_ind=Y What should I be doing instead?
[[alternative HTML version deleted]]

__
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[R] ROC Curves for Neural Networks

2013-06-07 Thread Richard Beckett 
Trying to plot an ROC curve to determine effectiveness of a neural network 
model in R. 

using the neuralnet package I created nn1a: 
nn1a-neuralnet(warstns ~ aigenlz+ 
p2l+anoclinv2+flinstab+mill+minconl+fllgdppclz+lpoplz+floil+flmtnestz+c_peaceyears,
 data=na.omit(dfram1), hidden=4) 

then using the lroc function from the epicalc package I tried lroc(nn1a) but 
came up with the error message 

Error in `colnames-`(`*tmp*`, value = c(Non-diseased, Diseased)) : 
  length of 'dimnames' [2] not equal to array extent 

Does anyone know any ways to get ROC curves to work with neural network models? 
Any other packages, model manipulation, etc? 

[[alternative HTML version deleted]]

__
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

Re: [R] Clogit R and Stata

2013-06-07 Thread Richard Beckett 





 From: peter dalgaard pda...@gmail.com

Cc: r-help@r-project.org r-help@r-project.org 
Sent: Friday, June 7, 2013 11:12 AM
Subject: Re: [R] Clogit R and Stata


Here is the R output:

Call:
coxph(formula = Surv(rep(1, 1404L), sftpcons) ~ sftptv2a3 + sftptv2a4 + 
    sftptv2a5 + sftptv2a2 + sftptv2a6 + logim + maccat + disp4cat + 
    strata(stratida), data = dframe, method = exact)

  n= 1404, number of events= 351 

             coef           exp(coef) se(coef)      z Pr(|z|)  
  
sftptv2a3  1.4552    4.2852   0.2273  6.401 1.54e-10 ***
sftptv2a4  3.1118   22.4609   0.2265 13.739   2e-16 ***
sftptv2a5  1.0717    2.9204   0.2522  4.249 2.15e-05 ***
sftptv2a2  0.7185    2.0514   0.3300  2.177   0.0295 *  
sftptv2a6  2.7341   15.3965   0.5050  5.414 6.17e-08 ***
logim      0.7579    2.1338   0.1347  5.625 1.85e-08 ***
maccat     3.0809   21.7771   0.4005  7.693 1.43e-14 ***
disp4cat   0.7061    2.0261   0.1524  4.634 3.59e-06 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ 
’ 1 

                   exp(coef) exp(-coef) lower .95 upper .95
sftptv2a3     4.285    0.23336     2.745     6.691
sftptv2a4    22.461    0.04452    14.409    35.013
sftptv2a5     2.920    0.34241     1.781     4.788
sftptv2a2     2.051    0.48747     1.074     3.917
sftptv2a6    15.397    0.06495     5.722    41.429
logim         2.134    0.46866     1.639     2.779
maccat       21.777    0.04592     9.934    47.739
disp4cat      2.026    0.49355     1.503     2.731

Rsquare= 0.239   (max possible= 0.623 )
Likelihood ratio test= 383.2  on 8 df,   p=0
Wald test            = 264.7  on 8 df,   p=0
Score (logrank) test = 396.2  on 8 df,   p=0


And the STATA output:

Iteration 0:   log likelihood = -95.537697  
Iteration 1:   log likelihood = -91.465581  
Iteration 2:   log likelihood = -91.402366  
Iteration 3:   log likelihood = -91.402264  
Iteration 4:   log likelihood = -91.402264  

Conditional (fixed-effects) logistic regression   Number of obs   =       
 468
                                                  LR 
chi2(8)      =     141.59
                                                  Prob 
 chi2     =     0.
Log likelihood = -91.402264                       Pseudo R2       
=     0.4365

--
    sftpcons |      Coef.   Std. Err.      z    P|z|     [95% 
Conf. Interval]
-+
   sftptv2a3 |   2.042827   .4741327     4.31   0.000     1.113544   
  2.97211
   sftptv2a4 |    4.10828   .5593723     7.34   0.000      3.01193  
  5.204629
   sftptv2a5 |   1.766492   .5585173     3.16   0.002     .6718177   
 2.861165
   sftptv2a2 |   1.366568   .6540307     2.09   0.037      .084691   
 2.648444
   sftptv2a6 |   2.307152   .8225835     2.80   0.005     .6949178   
 3.919386
       logim |   1.404135   .3480976     4.03   0.000     .7218764  
  2.086394
      maccat |     2.8423   .7008588     4.06   0.000     1.468642  
  4.215958
    disp4cat |   .6347805   .2872258     2.21   0.027     .0718283   
 1.197733
--

Also tried changing method=approximate with no noticeable change 
On Jun 7, 2013, at 15:34 , Richard Beckett wrote:

 Sorry to once again write a message but I'm once again stumped and am having 
 no luck finding a solution anywhere else.
 
 
 This question requires some finesse in both R and STATA so hopefully I will 
 be able to get an answer here. I am much more adept in R and am trying to 
 replicate the results of a STATA file in R. Hopefully this is a proper forum 
 for such questions. 
 
 
 This is the code for the clogit in STATA
 clogit sftpcons sftptv2a3 sftptv2a4 sftptv2a5 sftptv2a2 sftptv2a6 logim 
 maccat disp4cat if sample==1  glb_ind==Y, group(stratida)
 and I tried to replicate it using
 clogit1-clogit(sftpcons~sftptv2a3+sftptv2a4+sftptv2a5+sftptv2a2+sftptv2a6+logim+maccat+disp4cat+strata(stratida),
  dframe, sample==1 | glb_ind==Y)
 but got different results
 What did I do wrong here? I interpreted the STATA clogit as run this logit as 
 long as the sample is 1 and glb_ind=Y What should I be doing instead?


An  rather than | in the R version might help. Other than that, we're a 
bit short on clues unless you provide some output.

-- 
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd@cbs.dk  Priv: pda...@gmail.com
[[alternative HTML version