similar to: perturb package for evaluating collinearity

Dear Users, I try to calculate condition indexes and variance decomposition proportions in order to test for collinearity using colldiag() in perturb package, I got a large index and two variables with large variance decomposition proportions,but one of them is constant item.I also checked the VIF for that variable, the value is small.The result is as follows: Index intercept V1

Hello, could anyone explain what am I doing wrong. When I use colldiag function from package perturb I get different Variance Decomposition Proportions matrix in R than in SAS, although the eigenvalues and indexes are the same. Thanks for your attention. Results: in R: eigen(cor(indep2)) $values [1] 4.197131e+00 6.674837e-01 9.462858e-02 4.070314e-02 5.323022e-05 colldiag(indep2,c=T)

I have a linear model y~x1+x2 of some data where the coefficient for x1 is higher than I would have expected from theory (0.7 vs 0.88) I wondered whether this would be an artifact due to x1 and x2 being correlated despite that the variance inflation factor is not too high (1.065): I used perturbation analysis to evaluate collinearity library(perturb)

Hi everyone, I'm trying to vectorize an application of lda to each 2D slice of a 3D array, but am running into trouble: It seems there are quite a few 2D slices that trigger either the "variables are collinear" warning, or worse, trigger a "variable appears to be constant within groups" error and fails (i.e., ceases computation rather than skips bad slice). There are

You are using a method that needs to estimate the covariance matrix of all the variables. If you have 80 variables, there are (80+1)*80/2 = 3240 variances and covariances to estimate. How many data points do you think you need to do that? Some people assume the covariance matrix is diagonal (i.e., assuming all the variables are uncorrelated). Even then you still have 80 variances to estimate.

Hi, I am trying to find out a collinearity in explanatory variables with alias(). I creat a dataframe: dat <- ds[,sapply(ds,nlevels)>=2] dat$Y <- Response Explanatory variables are factor and response is continuous random variable. When I run a regression, I have the following error: fit <- aov( Y ~ . , data = dat) Error in "contrasts<-"(`*tmp*`, value =

is there any existing function for computing condition index? " analysing multivariate data" say that we can use condition index to check multicollinearity.saying that we can get it via SVD. The elements of the diagnoal matrix are the standard deviations of the uncorrelated vectors. the condition index is the ratio of the largest of these numbers to the smallest. so if i have a data

Dear all, How can I test for collinearity in the predictor data set for multiple linear regression. Thanks Alex [[alternative HTML version deleted]]

I am running a logistic regression model with a random effect using lmer. I am uncertain how to check for collinearity between my parameters. I have already run cor() and linear regression for each combination of parameters, and all Rsqr values were <0.8….but I am analyzing ecological data so a 0.8 cutoff may be unrealistic. -is there a way to check variance inflation factors or tolerance

A multinomial logit model can be specified as a conditional logit model after restructuring the data. Doing so gives flexibility in imposing restrictions on the dependent variable. One application is to specify a loglinear model for square tables, e.g. quasi-symmetry or quasi-independence, as a multinomial logit model with covariates. Further details on this technique and examples with several

Hi, and thanks for your help in order to do collinearity analysis I downloaded the perturb package. I run a lm (regression) and on that the ??calldiag?? commad to get condition numbers but i get the following message: the variable XY with modus ??numeric?? was not found (it does the same with all predictors despite all variables are numeric and exists). Can anyone tell me how can I go arround

Dear R users, Usually the variance-inflation factor, which is based on R^2, is used as a measure for multicollinearity. But, in contrast to OLS regression there is no robust R^2 available for MM-regressions in R. Do you know if an equivalent or an alternative nmeasure of multicollinearity is available for MM-regression in R? With best regards, Carsten Colombier Dr. Carsten Colombier Economist

I'm writing a program to perform linear regressions to estimate the number of bank teller transactions per hour of various types based upon day of week, time of day, week of month and several prices. I've got about 25,000 records in my dataset, 85 columns of transaction counts (used 1 at a time), about 50 columns of binary indicators (day, week, pay period, hour, branch), and a half dozen

Dear R-help list, apparently lda from the MASS package can be used in situations with collinear variables. It only produces a warning then but at least it defines a classification rule and produces results. However, I can't find on the help page how exactly it does this. I have a suspicion (it may look at the hyperplane containing the class means, using some kind of default/trivial

hello list, when I use method lda of the MASS package I experience a warning: variables are collinear in: lda.default(data[train, ], classes[train]) Is there an easy way to recover from this issue within the MASS package? Or how can I tell how severe this issue is at all? I understand that I shouldn't use lda at all with collinear data and should use "quadratische" (squared?)

Full_Name: John Hendrickx Version: 1.6.0 OS: Windows 98 Submission from: (NULL) (137.224.174.216) CrossTable in the "gregmisc" package fails when the fisher.exact test produces an error (I suspect this is because the number of cases is too large). This can be fixed using "FTt <- try(fisher.test(t, alternative = "two.sided"))" or by making the test optional.

R has amazing capabilities, but percentage tables are a weak spot IMHO. There's prop.table but that's rather unwieldly, especially for multiway tables. CrossTable by Marc Schwartz in the gregmisc library makes percentage tables a breeze but is limited to two-way tables. So I decided to try my own hand at writing an R-function that would make it easy to produce nicely formatted percentage

I have two specific questions regarding the output of lda function in MASS. #Question1: #========= n: sample size, p: number of variables Some articles in the literature say that LDA is singular for p > n-1. However, my experimentation with lda (default arguments) for two class problems shows collinearity for p > n-2. Does anyone know why this is the case? Does lda (MASS) use a different

I am using the Akima interpolation package to generate an interpolated color contour plot. It is working very well, except for one problem. The data that I have represents real-time readings from a thermistor string vs. time, so the data points are often very nearly in a rectangular array, since the thermistors are read at regular time intervals and they are equally spaced physically.

Is there a colinearty function implemented in R? I have tried help.search("colinearity") and help.search("collinearity") and have searched for "colinearity" and "collinearity" on http://www.rpad.org/Rpad/Rpad-refcard.pdf but with no success. Many thanks in advance, Peter Lauren.