Re: [R] need help with an interaction term

/LR_Mesh.htm but could not yield any plausible results. Hopefully this time... best regards Christian > -Ursprüngliche Nachricht- > Von: Chuck Cleland <[EMAIL PROTECTED]> > Gesendet: 31.08.06 14:16:12 > An: Christian Jones <[EMAIL PROTECTED]> > CC: r-help@stat.mat

[R] need help with an interaction term

Hello! IŽm fitting a model with glm(family binomial). The best model counts 9 Variables and includes an interaction term that was generated by the product of to continuous variables (a*b). All variables are correlated under a value of 0.7 (Spearman rank order) While the estimates of both main ef

[R] fitting an interaction term

Hello! IŽm fitting a model with glm(family binomial). The best model counts 9 Variables and includes an interaction term that was generated by the product of to continuous variables (a*b). All variables are correlated under a value of 0.7 (Spearman rank order) While the estimates of both main ef

[R] hier.part function???

can it also be, that a high joint contribution means, that the relevant variables only show there full effects on the response Variable in combination with other parameters in the model? Many thanks Christian Jones XXL-Speicher, PC-Virenschutz, Spartarife & mehr: Nur im WEB.DE

[R] problems with bootstrapping (Divergence or singularity ???)

while bootstrapping my fitted model for an internal validation: fit<-lrm(y~a+b+c ,x=T,y=T) val<-validate.lrm(fit,B=300,bw=T,rule="aic") R brings the notice that in some cases of the iterative process no variable has been choosen for estimation("Divergence or singularity in 129 samples"). I c

[R] predicting glm on a new dataset

Hello together, I would like to predict my fitted values on a new dataset. The original dataset consists of the variable a and b (data.frame(a,b)). The dataset for prediction consists of the same variables, but variable b has a constant value (x) added towards it (data.frame (a,b+x). The pre

[R] standardizing data

Hello R team, IŽm looking for a way to standardize (z transformation= standard deviation 1 and mean 0) a row of x y coordinates in order to conduct a trend analysis. Does anyone know the command in R? many thanks for help in advance Christian __ R-hel

[R] combining variables with PCA

hello R_team having perfomed a PCA on my fitted model with the function: data<- na.omit(dataset) data.pca<-prcomp(data,scale =TRUE), I´ve decided to aggregate two variables that are highly correlated. My first question is: How can I combine the two variables into one new predictor? and seco

[R] FW: aggregating variables with pca

Christian Jones <[EMAIL PROTECTED]> schrieb am 19.01.06 16:58:58: hello R_team having perfomed a PCA on my fitted model with the function: data<- na.omit(dataset) data.pca<-prcomp(data,scale =TRUE), I´ve decided to aggregate two variables that are highly correlated. My first

[R] aggregating variables with pca

hello R_team having perfomed a PCA on my fitted model with the function: data<- na.omit(dataset) data.pca<-prcomp(data,scale =TRUE), I´ve decided to aggregate two variables that are highly correlated. My first question is: How can I combine the two variables into one new predictor? and sec

[R] extracting values with $

Hello R friends! I´ve come acooss two problems during my work 1.) I would like to extract only certain values (such as R2 and C ) from the output of several models based on a logistic regression modela<-lrm(y~x1+x2+x3) , modelb<-lrm(y~x2+x5+x9)... > modela$coef #works fine, not so > mod

[R] Problems with BIC (Bayesian Information Criterion)

Hi, I would like to compare several Generalized Linear Models on the basis of BIC. My models have a binary response variable and are fitted with the glm function. AIC works well, not so BIC I tried: testBIC<-glm(y~x1+x2+x3,binomial) > BIC(testBIC) Error in log(x) : Non-numeric argument to math

[R] having scaling problems with a histogram

Hello, I would like to create a histogram from a data collumn consisting of 4 classes (0; 0.05;0.5;25;75). Due to the difference in scale the classes 0;0.05 and 0.5 are displayed within one combined bin by default with the code:Hist(x, scale="percent", breaks="Sturges"). How can I display the

[R] generating response curves

Hello does anyone know how to visualize a response curve based on a regression model with lines rather than dots. Having a large number of parameters the following formula is to time consuming. Perhaps a built in function exists to speed up the process. Model1<-a~b #Setting the scale extent

[R] keeping interaction terms

Hello, while doing my thesis in habitat modelling I´ve come across a problem with interaction terms. My question concerns the usage of interaction terms for linear regression modelling with R. If an interaction-term (predictor) is chosen for a multiple model, then, according to Crawley its si