similar to: Convergence problem in gam(mgcv)

Dear listers, I'm using gam(from mgcv) for semi-parametric regression on small and noisy datasets(10 to 200 observations), and facing a problem of overfitting. According to the book(Simon N. Wood / Generalized Additive Models: An Introduction with R), it is suggested to avoid overfitting by inflating the effective degrees of freedom in GCV evaluation with increased "gamma"

I am using gam in mgcv (1.3-22) and trying to use gcv to help with model selection. However, I'm a little confused by the process of assessing GCV scores based on their magnitude (or on relative changes in magnitude). Differences in GCV scores often seem "obvious" with my poisson gams but with negative binomial, the decision seems less clear. My data represent a similar pattern as

I have 11 possible predictor variables and use them to model quite a few target variables. In search for a consistent manner and possibly non-manual manner to identify the significant predictor vars out of the eleven I thought the option "select=T" might do. Example: (here only 4 pedictors) first is vanilla with "select=F" >

Hello, I have a question regarding model selection and dropping of terms for GAMs fitted with package mgcv. I am following the approach suggested in Wood (2001), Wood and Augustin (2002). I fitted a saturated model, and I find from the plots that for two of the covariates, 1. The confidence interval includes 0 almost everywhere 2. The degrees of freedom are NOT close to 1 3. The partial

Hi, I am using GAMs (package mgcv) to smooth event rates in a penalized regression setting and I was wondering if/how one can select the order of the derivative penalty. For my particular problem the order of the penalty (parameter "m" inside the "s" terms of the formula argument) appears to have a larger effect on the AIC/deviance of the estimated model than the

Qusetion #1 ********* Model selection in GAM can be done by using: 1. step.gam {gam} : A directional stepwise search 2. gam {mgcv} : Smoothness estimation using GCV or UBRE/AIC criterion Suppose my model starts with a additive model (linear part + spline part). Using gam() {mgcv} i got estimated degrees of freedom(edf) for the smoothing splines. Now I want to use the functional form of my model

hi Using a ts or tprs basis, I expected gcv to decrease when increasing the basis dimension, as I thought this would minimise gcv over a larger subspace. But gcv increased. Here's an example. thanks for any comments. greg #simulate some data set.seed(0) x1<-runif(500) x2<-rnorm(500) x3<-rpois(500,3) d<-runif(500) linp<--1+x1+0.5*x2+0.3*exp(-2*d)*sin(10*d)*x3

What does Generalized Cross Validation score mean. I preform and ordisurf on an ordination (nmds) with an environmental variable. I am trying to figure out "how well" the environmental varibles predict/explain the sites placements in species space. Any help would be greatly appreciated. Any pointers to literature... would be welcome. thanks in advance, -- Stephen Sefick Research

Hello everyone. I am doing a logistic gam (package mgcv) on a pretty large dataframe (130.000 cases with 100 variables). Because of that, the gam is fitted on a random subset of 10000. Now when I want to predict the values for the rest of the data, I get the following error: > gam.basis_alleakti.1.pr=predict(gam.basis_alleakti.1, +

Hi R-users, I'm using R 2.11.1, mgcv 1.6-2 to fit a generalized additive mixed model. I'm new to this package...and just got more and more problems... 1. Can I include correlation and/or random effect into gam( ) also? or only gamm( ) could be used? 2. I want to estimate the smoothing function s(x) under each level of treatment. i.e. different s(x) in each level of treatment. shall I

In Dr. Wood's book on GAM, he suggests in section 4.1.6 that it might be useful to shrink a single smooth by adding S=S+epsilon*I to the penalty matrix S. The context was the need to be able to shrink the term to zero if appropriate. I'd like to do this in order to shrink the coefficients towards zero (irrespective of the penalty for "wiggliness") - but not necessarily all the

Hi, there, I have 5 datasets. I would like to choose a basis spline with same knots in GAM function in order to obtain same basis function for 5 datasets. Moreover, the basis spline is used to for an interaction of two covarites. I used "cr" in one covariate, but it can only smooth w.r.t 1 covariate. Can anyone give me some suggestion about how to choose a proper smoothing spline

For summary(GAM) in the mgcv package smooth the degrees of freedom for the F value for test of smooth terms are the rank of covariance matrix of \hat{beta} and the residuals df. I've noticed that in a lot of GAMs I've fit the rank of the covariance turns out to be 9. In Simon Wood's book, the rank of covariance matrix is usually either 9 or 99 (pages 239-230 and 259). Can anyone

Hi R-users I am trying to use the GAM function of the mgcv package. But I am having problem trying to specify the k parameter. Although I managed to run some models by giving to the parameter some (random) value, and it is explained by Wood (2006) that it does not seem to "really" affect the final result, I would like to grasp better its meaning. I understand that is the

Dear Mr. Simon Wood, dear list members, I am trying to fit a similar model with gam from mgcv compared to what I did with BayesX, and have discovered the relatively new possibility of incorporating user-defined matrices for quadratic penalties on parametric terms using the "paraPen" argument. This was really a very good idea! However, I would like to constraint the coefficients

Sometimes, for specific models, I get this error from predict.gam in library mgcv: Error in complete.cases(object) : negative length vectors are not allowed Here's an example: model.calibrate <- gam(meansalesw ~ s(tscore,bs="cs",k=4), data=toplot, weights=weight, gam.method="perf.magic") > test <- predict(model.calibrate,newdata) Error in

Hello useRs! I have % cover data for different plant species in 300 plots, and I use the ARCSINE transformation (to deal with % cover data). When I use a GLM I do not have any problem. But when I am trying to use a GAM model using mgcv package, to account for non-linearity I get an ?error message?. I use the following model: sp1.gam<-gam(asin(sqrt(0.01*SP1COVER))~

Dear all, I'm using the mgcv library by Simon Wood to fit gam models with interactions and I have been reading (and running) the "factor 'by' variable example" given on the gam.models help page (see below, output from the two first models b, and b1). The example explains that both b and b1 fits are similar: "note that the preceding fit (here b) is the same as

I received an assignment that I have to do in R, but I'm absolutely not very good at it. The task is the following: http://www.nabble.com/file/p22804957/question8.jpg To do this, we also get the following pieces of code (not in correct order): http://www.nabble.com/file/p22804957/hints.jpg I'm terrible at this and I'm completely stuck. The model I chose can be found in here:

Dear all, When using GAM function in R, we need to specify the degree of freedom for the smooth function (i.e. s=(x, df=#)). I am wondering how to choose an appropriate df. Thanks a lot, Jin ---- North Carolina State University USA --------------------------------- [[alternative HTML version deleted]]