similar to: Condition indexes and variance inflation factors

Hi, I recently ran into an inconsistency in the way model.matrix.default handles factor encoding for higher level interactions with categorical variables when the full hierarchy of effects is not present. Depending on which lower level interactions are specified, the factor encoding changes for a higher level interaction. Consider the following minimal reproducible example: -------------- >

Hello Tyler, You write that you understand what I am saying. However, I am now at loss about what exactly is the problem with the behavior of R. Here is a script which reproduces your experiments with three variables (excluding the full model): m=expand.grid(X1=c(1,-1),X2=c(1,-1),X3=c("A","B","C")) model.matrix(~(X1+X2+X3)^3-X1:X3,data=m)

Hello Tyler, I rephrase my previous mail, as follows: In your example, T_i = X1:X2:X3. Let F_j = X3. (The numerical variables X1 and X2 are not encoded at all.) Then T_{i(j)} = X1:X2, which in the example is dropped from the model. Hence the X3 in T_i must be encoded by dummy variables, as indeed it is. Arie On Thu, Nov 2, 2017 at 4:11 PM, Tyler <tylermw at gmail.com> wrote: > Hi

Hello Tyler, I want to bring to your attention the following document: "What happens if you omit the main effect in a regression model with an interaction?" (https://stats.idre.ucla.edu/stata/faq/what-happens-if-you-omit-the-main-effect-in-a-regression-model-with-an-interaction). This gives a useful review of the problem. Your example is Case 2: a continuous and a categorical regressor.

Hello Tyler, Thank you for searching for, and finding, the basic description of the behavior of R in this matter. I think your example is in agreement with the book. But let me first note the following. You write: "F_j refers to a factor (variable) in a model and not a categorical factor". However: "a factor is a vector object used to specify a discrete classification"

My apologies, I forgot to CC: to the list on my previous communication with Daniel. Jesse ----- Forwarded by Jesse Albert Canchola/EMVL/DIAG/US/BAYER on 08/21/2006 10:50 AM ----- Jesse Albert Canchola/EMVL/DIAG/US/BAYER 08/21/2006 09:36 AM To "Daniel Nordlund" <res90sx5 at verizon.net> cc Subject RE: [R] Permutations with replacement Thanks, Daniel. I need to enumerate

Hello, just to follow up a question from last week. Here what I've done so far (here an example): library(MCMCpack) Y=c(15,14,23,18,19,9,19,13) X1=c(0.2,0.6,0.45,0.27,0.6,0.14,0.1,0.52) X2a=c(17,22,21,18,19,25,8,19) X2b=c(22,22,29,34,19,26,17,22) X2 <- function()runif(length(X2a), X2a, X2b) model1 <- MCMCregress(Y~X1+X2()) summary(model1) but I am not sure if my X2-function is

The double quotes are required by SQL if a name is not of the form letter-followed-by-any-number-of-letters-or-numbers or if the name is a SQL keyword like 'where' or 'select'. If you are doing this from a function, you may as well quote all the names. -Bill On Fri, Oct 2, 2020 at 6:18 PM Philip <herd_dog at cox.net> wrote: > The \?2B\? worked. Have no idea why. Can

I?m trying to pull data from one table (batting) in the Lahman Baseball database. Notice X2B for doubles and X3B for triples ? fourth and fifth from the right. The dbGetQuery function runs fine when I leave there two out but I get error messages (in red) when I include 2B/3B or X2B/X3B. Can anyone give me some direction? Thanks, Philip Heinrich

If I remove all the export and change all the ' in ", it does work on SCO 3.2v5.0.4 |--------+-----------------------------> | | Roumen Petrov | | | <Roumen.Petrov at skal| | | asoft.com> | | | | | | 21/12/00 13:10 | | | |

Hi Philip, You've probably realized by now that R doesn't like column names that start with a number. If you try to access an R-dataframe column named 2B or 3B with the familiar "$" notation, you'll get an error: > library(DBI) > library(RSQLite) > con2 <- dbConnect(SQLite(), "~/R_Dir/lahmansbaseballdb.sqlite") > Hack12Batting <-

This is really a feature of SQL, not R. SQL requires that you double quote column names that start with numbers, include spaces, etc., or that are SQL key words. E.g., > d <- data.frame(Order=c("sit","stay","heel"), Where=c("here","there","there"), From=c("me","me","you")) >

Hello UseRs, I've got 3 variables, the dependent variable Y as well as a max and a min value of the independent variable (Xa and Xb) where in some cases Xa=Xb (so actually a single value for X). First I'd like to perform a regression, but my problem is that my X is a range (acutally a censored independent variable Xa-Xb) rather then one single value. I know already some possible

Q: What platform don't run this script: ---------------------------------------- #!/bin/sh export X0='x0' export X1a="$X0/1" export X1b='$X0/1' export X2a="$X1a/2" export X2b='$X1b/2' $SHELL <<EOF_2 $SHELL <<EOF_1 cat <<EOF #define a "$X2a/aa" #define b "$X2b/bb" EOF EOF_1 EOF_2

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

Hi I have a set of data with both quantitative and categorical predictors. After scaling of response variable, i looked for multicollinearity (VIF values) among the predictors and removed the predictors who were hinding some of the other significant predictors. I'm curious to know whether the predictors (who are not significant) while doing simple 'lm' will be involved in

Just to inform: I posted that before in R-sig-ecology but as it might be interesting also for other useRs, I post it also to the general r-user list: Hello Alexandre, thank you very much. I also found another way to extract summarizing information from lm results over e.g. 1000 repeated model runs: results2 <- t(as.data.frame(results)) summary(results2) Although some questions popped up in

Hi Arie, Given the heuristic, in all of my examples with a missing two-factor interaction the three-factor interaction should be coded with dummy variables. In reality, it is encoded by dummy variables only when the numeric:numeric interaction is missing, and by contrasts for the other two. The heuristic does not specify separate behavior for numeric vs categorical factors (When the author of

I think it is not a bug. It is a general property of interactions. This property is best observed if all variables are factors (qualitative). For example, you have three variables (factors). You ask for as many interactions as possible, except an interaction term between two particular variables. When this interaction is not a constant, it is different for different values of the remaining

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.