similar to: What is the most useful way to detect nonlinearity in lo

It is easy to spot response nonlinearity in normal linear models using plot(something.lm). However plot(something.glm) produces artifactual peculiarities since the diagnostic residuals are constrained by the fact that y can only take values 0 or 1. What do R users find most useful in checking the linearity assumption of logistic regression (i.e. log-odds =a+bx)? Patrick Foley patfoley at

The locfit library is now available through CRAN, in the Contributed R Code directory. Locfit fits local regression, likelihood and density estimation models, in the spirit of loess but with many additional features. To install, unpack the locfit_19980309.tar.gz file, and R INSTALL locfit Most of the functionality and examples on my home page http://cm.bell-labs.com/stat/project/locfit/ should

The locfit library is now available through CRAN, in the Contributed R Code directory. Locfit fits local regression, likelihood and density estimation models, in the spirit of loess but with many additional features. To install, unpack the locfit_19980309.tar.gz file, and R INSTALL locfit Most of the functionality and examples on my home page http://cm.bell-labs.com/stat/project/locfit/ should

A sample from a continuous distribution most likely has no meaningful sample mode. Shown below is a way to estimate the location of mode from a kernel density estimate (using the locfit package): > x <- rnorm(50) > library(locfit) Locfit for R. August 3, 2000. (Updated for R 1.7.0, March 21, 2003) Attaching package 'locfit': The following object(s) are masked from

To estimate the covariance matrix of e you could use the sample covariance matrix of the residuals. If desired, use its cholesky decomposition to transform to make the error approximately uncorrelated, then refit (and back-transform the coefficient matrix). Stacking the columns of Y and replicating X won't do what you write; it forces each univariate regression to have the same coefficients.

Dear R-devel, I'm working on Prof. Loader's new version of locfit to try to get it pass R CMD check. I'm almost there, but I have a problem with some Rd files that I hope some one can help me resolve. Here's an example: In the package there's a function called locfit.censor(). This function can be used in a few different ways: locfit.censor(x, y, cens, ...)

I was wondering if someone could help me with a problem getting locfit running. I am running Digital Unix 3.2D on an alphastation, with R 0.61.1 I downloaded the CRAN distribution of locfit. I had no problems installing it with R INSTALL (I installed it into my library rather than the default base library; the output is attached at the end of this email message). I then tried to test it with the

What is the way to add an arbitrary fit from a model to a lattice conditioning plot ? For example xyplot(v1 ~v2 | v3,data=mydata, panel=function(...){ panel.xyplot(...) panel.loess(...,col.line="red") } ) Will add a loess smoother. Instead, I want to put a fit from lm (but not a simple straight line) and the fit has to be done for each panel

The version of locfit on the web site mentioned apparently has been revised by Prof. Loader, and is newer than the CRAN version that I have been maintaining. If Prof. Loader is OK with it, I will take a look and see if I can get the new version into CRAN-conforming form and upload to CRAN. Meanwhile, make sure you're using the package from Prof. Loader's web page, instead of the one on

Hi, I finally managed to make it work (I just needed to have a X11 window open). Thank you very much for your help. Agnes -----Original Message----- From: Ted Harding [mailto:Ted.Harding at nessie.mcc.ac.uk] Sent: Friday, July 21, 2006 3:34 PM To: Paquet, Agnes Cc: r-help at stat.math.ethz.ch Subject: RE: [R] connection to X11 problem On 21-Jul-06 Paquet, Agnes wrote: > Dear List, >

Hello!! At first thank you all for your suggestions and your help. But I would like to get another group of datas from the density estimation. I don't know if these information are provided by locfit or I should use an other function like plot(). Description: Each number between 1 and 100 (only integer) for example should have only one density value. But I also would like to have the density

Hi, Our firm is looking to see if it is possible to wrap some of more advanced lib in R like modreg.dll and specifically Loess function in there for both regression and prediction into MSVC++. We will probably take it to either VB or Python for implementation issues. Given that locfit has many utitlities built in for data transformation, this would be a good thing. If that's not

Apologies for the coming to this late... 1. By now I hope Somkiat has realized that R-bugs is not the place to report problems in contributed packages. Please direct such reports to the package maintainer. 2. This is really user error. predict() expect the newdata to be a data frame containing variables with the same names as those used in the fitting process. E.g., you fitted the model with

Hello, Here is a problem I have had trying to install locfit, the package for R. I have already contacted Clive Loader who thinks the problem has more to do with R than locfit, so here is the point : The version of R I am currently running is R-base-0.90 that I installed using the R-base-0.90.1-1.i386.rpm package. I dowloaded the locfit package available at the CRAN site and ran : R INSTALL

Hello!!! Thank you all for your suggestion on my last question about locfit. Now I have another question: I would like to change the evalution structure ev in locfit() to a vector/matrix. I created a vector and a matrix: > vec <- mat.or.vec(101, 1) > vec <- c(0:100) > mat <- matrix(data = vec, nrow = 101, ncol = 1) and I tried to change the ev parameter: > fit <-

I have a dataset of pregnancy values for multiple years (and ages, not included) with missing years. I would like to use local polynomial regression to smooth the values and estimate for the missing years. I would also like to use GCV to justify the smoothing parameter selection. When using locfit() with lp() I found that the gcvplot function does not work as it is looking for an alpha value to

Full_Name: Somkiat Apipattanavis Version: 2.1.1 OS: Windows Submission from: (NULL) (128.138.44.123) Bug found in predict.locfit for density estimation # Example of bug found in prdict.locfit (Locfit) library('locfit') # generate data y =c(4281,2497,4346,5588,5593,3474,4291,2542,5195,4056, 3114,2864,4904,7625,3377,4001,4999,7191,8062,5668) x1=c( 0.258729, 1.460156, 0.192323,

Hi All, In another thread Andy Liaw, who CRAN lists as locfit maintainer; said: <quote> From: "Liaw, Andy" <andy_liaw at merck.com> To: "Guy Green" <guygreen at netvigator.com>; <r-help at r-project.org> Subject: Re: Alternatives to linear regression with multiple variables Date: 22 February 2010 17:50 You can try the locfit package, which I believe

How about: nis.na <- complete.cases --- From: <Ted.Harding at nessie.mcc.ac.uk> Hi Folks, My recent response to Laura Quinn's query about matrix subsetting reminded of a question. I wrote: iDir <- ((Winds[,20]<45)|(Winds[,20]>315))&(!is.na(Winds[,20])) Now, I find "!is.na" a bit awkward to type, so I might prefer to type it as "nis.na".

Dear list, I'm a bit perplexed why the 95% confidence bands for the predicted probabilities for units where x=0 and x=1 overlap in the following instance. I've simulated binary data to which I've then fitted a simple logistic regression model, with one covariate, and the coefficient on x is statistically significant at the 0.05 level. I've then used two different methods to