Hi folks, I have an S-PLUS program that I would like to call from a DOS prompt. I know I can call an S-PLUS program named program.ssc from DOS by typing: splus.exe program.ssc This works successfully
Hi folks, I have a data frame containing a column of type 'factor'. This column contains three distinct values, namely, 'Current', 'lastyr', and '2yrsago'. I would like to use the recode function fro
Hi everyone, Hopefully someone can help with this problem. I have a very large amount of data which I need to reduce in size for some statistical routines I wish to run. In order to keep the appropri
Thanks for the response Prof. Ripley. I'm guessing that I should create a new column containing an index variable for which I can stratify my data. I have several columns which represent characterist
Hi all, I am interested in evaluating the differences between logistic regression output in S-PLUS and SAS. The situation is this: I have a model of the following form: Y ~ X* where X* is a categoric
Many thanks to Frank Harrell Jr. and Tom Richards who gave the same helpful response to my query! In S+ just do options(contrasts=c('contr.treatment','contr.poly')). The first level of factor variabl
Hi all, I have a relatively simple graphing question I believe. Suppose I have a data frame that looks like this: V1 V2 -- abc .11 def .12 abc .18 def .24 abc .33 where V1 is a factor variable and V2
Hi all, Simple question this time folks. Does anyone have any experience with StatServer? If not, does anyone know of a news group devoted to StatServer? We use S-PLUS 2000 and StatServer here, and w
tion from the Mass library. I used this function many many times before without problems. This time after the function ran for a while, in fact I am pretty
to use an offset variable in a poisson regression (glm). Here are the details: I have a written a program in S-PLUS which runs the following glm call: glm
riables does not hurt. It actually hurts if you remove variables on the basis of P-values because of biases in estimating the residual variance and bias in
e difference is undoubtedly much smaller than the variability of these statistics! mean() divides by n, var() divides by n-1; I think you want to divide by
I am new to the S-news list, and I am hopeful that someone will be able to answer some of my questions. My background is mostly in SAS, so I am relatively new to S/S-PLUS. If these questions are rep
I did receive a couple responses to question 1 which were helpful. To parameterize the equation I ended up doing the following: rhsformula <- eval(parse(text = paste("Yresponse","~",paste(dims,colla
Hi everyone, I have a question about limits on the size of data frames in S-PLUS. I use S-PLUS 2000 on Windows NT 4.0, and have to use very large data sources. The computer is a Pentium III 600MHz pr
Thanks Win for your response! I thought I would forward it to everyone... Reid, I'm sure you will get a response from the S-PLUS developers, but I recently ingested a file, with no problems, that had
Hi all, I have a relatively simple question. How can one go about calculating correlations between variables when one or more of them are factors variables? Is there an alternative means to measure l
Hi all, I have read a lot of contrasting information regarding this topic, and I wanted to get some others' opinions. Oftentimes I need to fit regression models where my dependent variable is a rate.
Hi all, I have a column of character data in an S-PLUS data frame of the following form: substring.substring substring.substring substring.substring substring.substring etc. where substring is just p
Thanks Samuel, I'm afraid I didn't give enough information related to my problem. I do indeed need to locate the "." character in each string, and to further complicate matters, I may have more than