similar to: Resolution of problem re location polr and of MASS

Hello I am running R 1.8.1 on a Windows platform I am attempting to fit an ordinal logistic regression model, using the polr function, as described in Venables and Ripley. But when I try model4 <- polr(ypsxcat~committed + as.factor(sex) + as.factor(drugusey) + anycsw + as.factor(sex)*committed + as.factor(sex)*as.factor(drugusey)+as.factor(sex)*anycsw, data = duhray) I get a message

Hello again I am trying to fit an ordinal logistic model using the polr function from MASS. When I run model.loan.ordinal <- polr(loancat~age + sex + racgp + yrseduc + needlchg + gallery + sniffball + smokeball + sniffher + smokeher + nicocaine + inject + poly(year.of.int,3) + druginj + inj.years) I get an error Error in optim(start, fmin, gmin, method = "BFGS", hessian =

I'd like to thank John Fox and Chuck Cleland for their help in resovling this issue. It turned out to be something simple, but perhaps others have had similar problems In my original data frame, I had 4 categories of race/ethnicity. One of the categories (other) was very small, and not similar to any of the other three categories, so I created a new data frame deleting those people.

John Fox wrote <<< >From your description, it seems possible that there are too many zeros for a Poisson or negative-binomial model. Since the focus of your paper is the methodology, you might want to try a zero-inflated Poisson or negative-binomial model. Though I haven't tried them, I'm aware of two sources of R functions for zero-inflated count models -- zeroinfl(), from

Does anyone know of research on the effects of rounding on regression? e.g., when you ask people "How often have you _______?" you are more likely to get answers like 100, 200, etc. than 98, 203, etc. I'm interested in investigating this, but don't want to reinvent the wheel. thanks Peter Peter L. Flom, PhD Assistant Director, Statistics and Data Analysis Core Center for

Hello all I have a feeling this is very simple......but I am not sure how to do it My boss has two variables, one is an average of 4 numbers, the other is an average of 3 of those numbers i.e var1 = (X1 + X2 + X3 + X4)/4 var2 = (X1 + X2 + X3)/3 all of the X variables are supposed to be measuring similar constructs not surprisingly, these are highly correlated (r = .98), the question is how

Hello Using R 2.2.1 on a Windows machine. Has anyone programmed the Welch test for equality of variances? I tried RSiteSearch, but this gave references to t test and oneway.test, which are not quite what I need.....I need the Welch test itself, for use in a meta-analysis (to determine if variances are equal). TIA Peter Peter L. Flom, PhD Assistant Director, Statistics and Data Analysis

Thanks for all the help on my earlier questions..... How do you plot a simple time series with unequal intervals? I have the following dateofpoll <- as.ts(c("6/1/02", "7/1/02", "10/1/02", "1/4/03", "1/25/03", "6/7/03", "7/16/03", "8/17/03", "9/4/03")) reelect <- c(51, 47, 49, 51, 49, 49, 46, 45, 40)

Pursing my earlier question, when I tried loading Lindsey's gnlm, I got a message Loading required package: rmutil Warning message: There is no package called 'rmutil' in: library(package, character.only = TRUE, logical = TRUE, warn.conflicts = warn.conflicts, According to the R documentation http://finzi.psych.upenn.edu/R/doc/html/packages.html rmutil is in the standard

Hello I am running R 2.0.1 on Windows, I am attempting to use Frank Harrell's 'fastbw' function (from the Design library), but I get an error that the fit was not created with a Design library fitting function; yet when I go to the help for fastbw (and also look in Frank's book Regression Modeling Strategies) it appears that fastbw should work with a model created with lm.....

Running R 1.8.1 on a Windows machine In dotchart, I would like to shrink the labels on the tick marks (that is, the numbers) without shrinking anything else. I could not find this in either the Rhelp archives or in ?dotchart, which recmmends cex to avoid 'label overlap', but cex shrinks all the characters in the plot. Is there a way to do this? Thanks Peter Peter L. Flom, PhD

Sorry to keep asking elementary questions......I appreciate the help. I am trying to create a dotchart with the rows sorted according to the values, rather than the labels. When I try prof <- c('Accountant', 'Administrative assistant', 'Garment worker', 'Cook', 'Dentist', 'General practictioner', 'Graduate student', 'High level

When I try citypro <- sas.get('c:/ndri/cvar/data', member = 'citypro2') I get the following errors and warnings: 'sas' is not recognized as an internal or external command, operable program or batch file. 'less' is not recognized as an internal or external command, operable program or batch file. Error in sas.get("c:/ndri/cvar/data", member =

I was under the impression that the tree method (e.g. as implemented in rpart) was insensitive to monotonic transformations of the dependent variable. e.g. Breiman Olshen et al. Classification and Regression Trees state "In a standard data structure [a tree] is invariant under all monotone transformations of individual ordered varaibles" (p. 57) However, I get very different results

I have a dataframe called cvar, with two variables (among many others) called MSA and ACTUP. Both are numeric. This was working fine. Then I found out that for two MSAs, ACTUP should be 1, not 0. so I tried cvar$ACTUP[cvar$MSA == 6840] <- 1 cvar$ACTUP[cvar$MSA == 5360] <- 1 but when I try table(cvar$MSA, cvar$ACTUP) the level of ACTUP for those two MSAs has not changed, and is still

Hello When I run scatter.smooth(jitter(weight), jitter(height2), span = .25, evaluation = 50, pch = '.') I get the type of graph I thought I would get, but also a warning..... k-d tree limited by memory. ncmax= 528 I always get concerned when there are warnings I don't understand. What's a k-d tree? Is this something to be concerned about? Thanks Peter Peter L. Flom,

Hello I have a dataframe that looks like this MSA CITY HIVEST YEAR YR CAT 1 0200 Albuquerque 0.50 1996 1996 5 2 0520 Atlanta 13.00 1997 1997 5 3 0720 Baltimore 29.10 1994 1994 1 4 0720 Baltimore 13.00 1995 1995 5 5 0720 Baltimore 3.68

Hello Pardon for the elementary question, I did try searching the archives with various terms, but to no avail. I am using R1.9.1 on a windows machine One of the great advantages of R (to me, anyway) is being able to see the code for a function , e.g. by typing sd one sees the code for getting a standard deviation. However, for many functions this only provides info. including UseMethod, eg.

Hello again I have modeled a tree using rpart, with the DV being a log transformation of the variable I am really interested in (I transformed the DV due to extreme skewness). By default, text.rpart labels the nodes with the value of yval, which in this case is not what I want; I'd like the labels to be on the original metric, but label in text.rpart requires a "column name of

Hello I am interested in Poisson or (ideally) Negative Binomial regression with an inflated number of 1 responses I have seen JK Lindsey's fmr function in the gnlm library, which fits zero inflated Poisson (ZIP) or zero inflated negative binomial regression, but the help file states that for ' Poisson or related distributions the mixture involves the zero category'. I had thought