similar to: aic calculus for glm models

The model frame shows the response and predictors in a data frame with nicely labelled columns: fm <- lm(wt~qsec+log(hp)+sqrt(disp), data=mtcars) model.frame(fm) # ok When the left hand side consists of more than one response, those response variables still look good, inside a matrix: fm <- lm(cbind(qsec,hp,disp)~wt, data=mtcars) model.frame(fm)[[1]] # ok A problem arises when

I think you are not using the best function for what your intentions are. Try: > by(data=mtcars, INDICES=list(as.factor(mtcars$am)), FUN=colMeans) : 0 mpg cyl disp hp drat wt qsec vs 17.1473684 6.9473684 290.3789474 160.2631579 3.2863158 3.7688947 18.1831579 0.3684211 am gear carb 0.0000000

Hola, Bueno, puedes hacer el cálculo de una forma mucho más compacta y rápida. Esta forma es especialmente recomendable cuando tienes muchas columnas y muchas filas. > library(data.table) > myDT <- as.data.table(mtcars) > myDTlong <- melt(myDT, measure.vars=1:ncol(myDT)) > myDTlong[ , list(p_value = shapiro.test(value)$p.value, v_stat = shapiro.test(value)$statistic) , by

Dear list, I just hit an error that stopped my make && make check-devel operation on my linux box (FC3, i686 P4 2GB RAM). Just to note that I've been building the development branch(?) for some time and this is the first hint of a problem. 1) updated the src tree using svn update 2) ran ../configure --with-recommended-package=no from my build directory 3) got: R is now configured

Hi folks, I'm trying to estimate bias-corrected percentile (BCP) confidence intervals on a vector from a simple for loop used for resampling. I am attempting to follow steps in Manly, B. 1998. Randomization, bootstrap and monte carlo methods in biology. 2nd edition., p. 48. PDF of the approach/steps should be available here: https://wyocoopunit.box.com/s/9vm4vgmbx5h7um809bvg6u7wr392v6i9 If

Hello all, I was wondering if someone can enlighten me as to the difference between the logLik in R vis-a-vis Stata for a GLM model with the gamma family. Stata calculates the loglikelihood of the model as (in R notation) some equivalent function of -1/scale * sum(Y/mu+log(mu)+(scale-1)*log(Y)+log(scale)+scale*lgamma(1/scale)) where scale (or dispersion) = 1, Y = the response variable, and mu

Hi everyone, I have another "Random Forest" package question: - my (presumably incorrect) understanding of the varImpPlot is that it should plot the "% increase in MSE" and "IncNodePurity" exactly as can be found from the "importance" section of the model results. - However, the plot does not, in fact, match the "importance"

Dear All, Thanks for your help. However, I would like to draw your attention to the following: Actually, I was replicating the Example 2.3, using the dataset "brainsize.txt" given in Section 2.3.3 ("Summarize by group") at page 55, of a famous book "R by Example" written by "Jim Albert and Maria Rizzo" published in Springers (2012) in a Use R! Series. The

Dear Sirs, I am Professor at Indira Gandhi Krishi Vishwavidyalaya, Raipur, Chhattisgarh, India. While taking classes, I found the *by() *function producing following error when I use FUN=mean or median and some other functions, however, FUN=summary works. Given below is the output of the example I used on a built-in dataset "mtcars", along with error message reproduced herewith: >

> On Apr 15, 2016, at 1:16 AM, Akhilesh Singh <akhileshsingh.igkv at gmail.com> wrote: > > Dear All, > > Thanks for your help. However, I would like to draw your attention to the > following: > > Actually, I was replicating the Example 2.3, using the dataset > "brainsize.txt" given in Section 2.3.3 ("Summarize by group") at page 55, > of a

Dear All, I have got your core message, that it is my responsibility to determine whether any particular function in my version of R satisfies the language requirements at the time of your use. Jim Albert and Maria Rizzo must have used their code, which was permitted in the R-code of their time (2012). Therefore, I have now modified my R-code, as per R-3..2.4 version, according to my requirement

Hello, I have found a problem with using the glm function with a gamma family. I have a vector of data, assumed to be generated by a gamma distribution. The parameters of this gamma distribution are estimated in two ways (i) using the glm() function, (ii) "by hand", using the optim() function. I find that the -2*likelihood at the maximum found by (i) is substantially larger than that

> On Apr 16, 2016, at 2:03 AM, Akhilesh Singh <akhileshsingh.igkv at gmail.com> wrote: > > Dear All, > > I have got your core message, that it is my responsibility to determine whether any particular function in my version of R satisfies the language requirements at the time of your use. Jim Albert and Maria Rizzo must have used their code, which was permitted in the R-code

Dear R-List members, I want to check if a set of measurements follows better a gamma or a lognormal distribution (see data below). Using shapiro.test I can test for normality (shapiro.test(log (Lt)). To test for gamma (and normal) distribution I would use ks.test but I need to specify its shape and scale. How should I calculate these values in R? I tried > Lt.fit <- glm(Lt ~ 1,

I have applied the same linear model to several different subsets of a dataset. I recently read that in R, code should never be repeated. I feel my code as it currently stands has a lot of repetition, which could be condensed into fewer lines. I will use the mtcars dataset to replicate what I have done. My question is: how can I use fewer lines of code (for example using a for loop, a function or

Dear list, I want the 1st, 2nd, 5th, and 6th columns of mtcars. After copying them, the columns become numeric class rather than data frame. But, when I copy rows, they data frame retains its class. Why is this? I don't see why copying rows vs columns is so different. > class(mtcars) [1] "data.frame" > head(mtcars) mpg cyl disp hp drat wt qsec vs

Hola a tod en s, bueno, un poco al hilo de un mensaje anterior, reformulo la pregunta y pongo código completamente reproducible, a ver si hay más suerte. Estoy con un informe y hay que hacer algunos gráficos con datos diferentes con ggplot2, y hay que poner el mismo tamaño para todos ellos. El problema es que R cambia automáticamente el tamaño del área de trazado y el margen en función de

I have been successfully using RODBC for a long time (years) to connect to MS SQL Server from R. This week I wanted to try using odbc but I am seeing some problems which may be related to how I set up my driver and/or connection. The dbWriteTable manual page gives as an example command: dbWriteTable( pDB$con, "mtcars", mtcars[1:5,]) When I try this I get the following error Error:

Hello, Seeking a way to get the variance inflation factor VIF for a model of regression in complex survey, I have understood that without this package (SURVEY) RGui VIF obtained as follows: fit <- lm(mpg~disp+hp+wt+drat, data=mtcars) vif(fit) But I want to know if survey, Vif is obtained so vif( svyglm(api00~ell+meals+mobility, design=dstrat)) Thank you, happy day

Using the mtcars dataset, how to define the grouping variable to be valid only for the upper or lower panel? The following doesn't work: # Code start splom(~data.frame(mpg, disp, hp, drat, wt, qsec), data=mtcars, pscales=0, auto.key=list(columns=3), upper.panel = function(...){ panel.grid(...) panel.xyplot(groups=cyl,...) } ) # Code end -- View