similar to: cross-validation complex model AUC Nagelkerke R squared code

Dear R community. I´m working with a generalized linear model which the response variable is a categorical one and the predictive variables are weather conditions. I have 250 different places where I need to fit the model. In some of these places I have strong correlations between some of the variables so I need to deal with this problem. I found a work similar than mine where they use tha

Hello I found some small postings dated to 22 Oct 2008 on the message subject. Recently, I have been working with binary logistic regressions. I didn't use the design package. Yet, I needed the "fit" indices. Therefore, I wrote a small function to output the Nagelkerke's R, and the Cox-&-Snell R from a fitted model. I am no professional programmer by far, yet, I hope, that

Hello, When evaluating different learning methods for a categorization problem with the (really useful!) caret package, I'm getting confusing results from the Kappa computation. The data is about 20,000 rows and a few dozen columns, and the categories are quite asymmetrical, 4.1% in one category and 95.9% in the other. When I train a ctree model as: model <- train(dat.dts,

I am interested Andrea is whether you ever established why your R2 was 1. I have had a similar situation previously. My main issue though, which I'd be v grateful for advice on, is why I am obtaining such negative values -0.3 for Somers Dxy using validate.cph from the Design package given my value of Nagelkerke R2 is not so low 13.2%. I have this output when fitting 6 variables all with

You may recall that I was recently constructing a function to bootstrap the coefficients in a linear regression model. In S-PLUS I was using the model.matrix function applied to the fitted model, then taking the QR decomposition of that. I discovered that it was in fact easier to accomplish the bootstrapping in R because the QR decomposition of the model matrix is stored with the fitted model.

I am trying to show some techniques to my graduate regression class. The textbook mentioned using bootstrap samples of regression coefficients for assessing variability. I decided to show them reasonably effective ways of doing the resampling. The following is a function I wrote to create bootstrap samples of coefficients from a fitted linear regression model. bsCoefSample <- ##

Hello All, as I?m new to R and survival analysis, I?ve got a question about the Design::validate function: My Code: cox <- cph(Surv(t,status) ~ var1 + var2 + var3, data=data, x=TRUE, y=TRUE, surv=TRUE) cox.val <- validate(cox, B=10, dxy=TRUE, pr=TRUE); My output (cox.val): index.orig training test Dxy -0.3639222921368090891

When regressing by month, how do I get the coefficients out into a new data set? I'm looking for [ month, a, b, c ] from the Pastor-Stambaugh model I'm using which is: r[i+1] = a + b * r[i] + c * v[i] + e the model I'm using wants to create a new dataseries based on the coefficient in each month. I'm doing a simple linear regression on DataSet, and >

Hi I am trying to perform a bootstrap estimate of classification accuracy of a logistic regression using the 'Daim' package in r using the code at the bottom of this post, this all works great and I get the .632+ misclassification accuracy, specificity, sensitivity, AUC etc etc but what I would like is to access the list of AUC for each of the bootstrap samples as I need calculate the 95%

Hi there, I just don't find the solution on the following problem. :( Suppose I have a dataframe with two predictor variables (x1,x2) and one depend binary variable (y). How is it possible to define a glm object (family="binomial") with a user defined logistic function like p(y) = exp(a + c1*x1 + c2*x2) where c1,c2 are the coefficents which I define. So I would like to do no

Dear All,   I have the following example:   a <-matrix(c(1:100),ncol=10) b <-matrix(c(2,4,6,8,10,12,14,16,18,20))   trapz(b,a)   will give me a result of 99, which it seems to me is the AUC of the 1st column only. Is it possible to get the AUC results by columns of "a" using the same "b" values in the calculations as opposed to just generating the result for the 1st

Hi, I need to apply the Partial AUC in ROC curves. How I do this?. And I need the SE of partial AUC too. I have been reading ?caTool, ?Epi, ?verification, ?ROC, ?ROCR, ?genefilter and more. I only see the partial AUC, but I do not the SE. Thank you, if you can help me. Xavier G. Ordo?ez Doctoral Student

Hi Does anyone know whether the rms package provides a confidence interval for the bootstrap-corrected Dxy or c-index? I have fitted a logistic model, and would like to obtain the 95% confidence interval of the bootstrap-corrected area under the ROC curve estimate. Thanks. [[alternative HTML version deleted]]

Is there a statistical test implemented within R to compare the area under two ROC curves ?? Thanks, Berthold -------------- next part -------------- A non-text attachment was scrubbed... Name: berthold.kramm.vcf Type: text/x-vcard Size: 214 bytes Desc: Card for Dr. Berthold Kramm Url : https://stat.ethz.ch/pipermail/r-help/attachments/20011120/dc0f7bf1/berthold.kramm.vcf

I am using the rcorr.cens function from the Hmisc package and the time-dependent ROC curve obtained using tdrocc in the survcomp package. I understand that the C statistic from rcorr.cens has to be subtracted from 1 if high values of the risk variable lower survival. Given that I wonder what the connection is between that C statistic and the AUC from the tdrocc object. If they are substantially

Dear all, I am currently running an ENFA in R using Adehabitat pkg to assess species distribution and I am really stuck and need some help to finalize my project. I run the ENFA in both original location data and pseudo-absence points (the latter randomly generated using dismo pkg), and now I would like to assess the performance of the ENFA, however, I am not sure about how to tackle this. I

Hi, I have a dataset (see attached) with 2 variables "Y" is binary, "x" is a continuous variable. I want to calculate area under the curve (AUC) for the ROC curve, but I got different AUC values using ROC() from Epi package vs. rcorr.cens() from rms package: test<-read.table("test.txt",sep='\t',header=T,row.names=NULL) y<-test$y x<-test$x

Hello R-listers, I'm working in an experiment that try to determine the degree of infection of different clones of a fungus and, one of the measures we use to determine these degree is the counting of antibodies in the plasma at different dilutions, in this experiment the maximum number of dilutions was eleven. I already checked for differences on the maximum concentration of the

Here is a snippet to show what i'm trying to do. library(randomForest) library(ROCR) library(caret) data(iris) iris <- iris[(iris$Species != "setosa"),] fit <- randomForest(factor(Species) ~ ., data=iris, ntree=50) train.predict <- predict(fit,iris,type="prob")[,2]

Hi, there: Could someone tell me a simple function of plot ROC curve and calculate AUC in R? My setting is very simple, a column of the true binary response and another column of predicted probabilities. Thanks! Yulei [[alternative HTML version deleted]]