search for: influences

Hello, I am looking for an explanation and/or fix for a discrepancy in the behaviour of the R lm.influence() function [ version R 1.5.0 (2002-04-29) ] and the same function in Splus [ Splus version 5.1 release 1, running on SGI IRIX 6.2]. The discrepancy is of concern because I am migrating some Splus scripts to R and need to ensure consistency of results. Specifically, when I fit a glm()

>>>>> Murray Efford <murray.efford at otago.ac.nz> >>>>> on Fri, 8 Apr 2016 18:45:33 +0000 writes: > Thanks for these perfectly consistent replies - I didn't > understand the purpose of m = sum(w * f/sum(w)) and saw it > merely as a weighted average of the fitted values. My > ultimate concern is how to compute an appropriate

...efficients in R. This is supposed to be inv(information)*(y-yhat)*x. So I am wondering what is the exact mathematical formula for the output that the functions: dfbeta() OR lm.influence()$coefficients return for a glm model. I am confused because: 1. Their columns don't sum to zero as influences should. 2. They return different "influences", so the 2 functions are doing something different. 3. I think they divide each element by the standard error of the corresponding coefficient, but that's not enough to resolve any discrepancies The documentation doesn't provide any...

> On Apr 10, 2016, at 3:11 AM, Murray Efford <murray.efford at otago.ac.nz> wrote: > > Martin - > Thanks, but although hatvalues() is useful for calculating PRESS, I can't find anything directly relevant to my question in the influence help pages. After some burrowing in the literature I'm doubting there is an answer out there (PRESS R^2 is always presented in a fairly

Martin - Thanks, but although hatvalues() is useful for calculating PRESS, I can't find anything directly relevant to my question in the influence help pages. After some burrowing in the literature I'm doubting there is an answer out there (PRESS R^2 is always presented in a fairly ad hoc way). This is a new topic, as you say, and perhaps better handled on a statistics list. Murray Efford

Dear list, There is, I believe, a minor typo in the example section of influence.measures.Rd. In the final example the word `does` appears where I suspect `dose` is required: I couldn't remember exactly what format patches should be in, so here is one as diff would produce: Index: devel/src/library/stats/man/influence.measures.Rd

Hello everyone ! I am currently trying to convert a program from S-plus to R, and I am having some trouble with the S-plus function called "influence(data, statistic,...)". This function aims to "calculate empirical influence values and related quantities", and is part of the Resample library that I cannot find for R. However, 2 similar functions are available in R: - the

Hi all I'm getting a NaN returned on using dffits, as explained below. To me, there seems no obvious (or non-obvious reason for that matter) reason why a NaN appears. Before I start digging further, can anyone see why dffits might be failing? Is there a problem with the data? Consider: # Load data dep <-

I'm trying to write some functions extending influence measures to multivariate linear models and also allow subsets of size m>=1 to be considered for deletion diagnostics. I'd like these to work roughly parallel to those functions for the univariate lm where only single case deletion (m=1) diagnostics are considered. Corresponding to stats::hatvalues.lm, the S3 method for class

influence.measures gives several measures of influence for each observation (Cook's Distance, etc) and actually flags observations that it determines are influential by any of the measures. Looks good! But how does it discriminate between the influential and non- influential observations by each of the measures? Like does it do a Bonferroni-corrected t on the residuals identified by

Barrett & Ling, JASA, 1992, v.87(417), pp184-191 define general classes of influence measures for multivariate regression models, including analogs of Cook's D, Andrews & Pregibon COVRATIO, etc. As in univariate response models, these are based on leverage and residuals based on omitting one (or more) observations at a time and refitting, although, in the univariate case, the

> On Apr 10, 2016, at 9:38 AM, David Winsemius <dwinsemius at comcast.net> wrote: > >> >> On Apr 10, 2016, at 3:11 AM, Murray Efford <murray.efford at otago.ac.nz> wrote: >> >> Martin - >> Thanks, but although hatvalues() is useful for calculating PRESS, I can't find anything directly relevant to my question in the influence help pages. After

I'm following up on a question I posted 8/10/2010, but my newsreader has lost this thread. > Barrett & Ling, JASA, 1992, v.87(417), pp184-191 define general > classes of influence measures for multivariate > regression models, including analogs of Cook's D, Andrews & Pregibon > COVRATIO, etc. As in univariate > response models, these are based on leverage and

Dear list, I've noticed that influence.measures and cooks.distance gives different results for non-gaussian GLMs. For example, using R-1.9.0 alpha (2003-03-17) under Windows: > ## Dobson (1990) Page 93: Randomized Controlled Trial : > counts <- c(18,17,15,20,10,20,25,13,12) > outcome <- gl(3,1,9) > treatment <- gl(3,3) > glm.D93 <- glm(counts ~ outcome +

Hello everybody, a question that connect to the question of Frederik Karlsons about 'how to stand. betas' With the stand. betas i can compare the influence of the different explaning variables. What do i with the betas of factors? I can't use the solution of JohnFox, because there is no sd of an factor. How can i compare the influence of the factor with the influence of the numeric

Greetings! Recently when I tried to use lm.influence I get the following error: Error in .Fortran("lminfl", model$qr$qr, n, n, k, as.integer(do.coef), : Fortran symbol name "lminfl" not in DLL for package "base" This occurs on both Linux and Windows platforms (details below). Searching the mail lists and other sources indicates that the fortran code for

Hi all, Does lme has function to compute the cook's distance or influence measure like lm? I can't find them. Thanks. Yen Lin [[alternative HTML version deleted]]

Hi dear all, I'm wondering about the question that; Does the influence.measures(model) for linear models valid for general linear models such as logistic regression models? That is; If I fit the model like model <- glm( y~X1+X2, family=binomial) Then, if i apply the function "influence.measures(model), i will get the result of influence measures. These result are valid for

Dear R users, I'm trying to construct a distance matrix based on a nb object created by spdep where sites must have a larger influence in one direction then in the other. Here is an example to better illustrate what I need: Let say I have the following Gabriel connection network library(spdep) library(ade4) data(orbatid) nbgab<-graph2nb(gabrielneigh(as.matrix(oribatid$xy)))

I am a novice R user, and I am having difficulty understanding R's influence plots. I am trying to remove outliers from a particular variable, "sib." I am able to generate influence plots and further outlier information such as below (which is a shortened example). For my analyses, I end up excluding the points R refers to, 7, 18, 26, and 105. However, my question is, how can I