Re: [R] Request for aid in first R script

Dear Kepler, Yes, R can do this all. But this is is to help you when you get stuck, not to do all the work for you... You are asking basic stuff, so any introduction book on R should contain sufficient information to get you going. So please do read on of those first. Best regards, ir. Thierry O

Re: [R] R code for if-then-do code blocks

Dear Paul, R's power is that is works vectorised. Unlike SAS which is rowbased. Using R in a SAS way will lead to very slow code. Your examples can be written vectorised d1 %>% rownames_to_column("ID") %>% mutate( test1 = ifelse(gender == "f" & workshop == 1, 7, 0), test2 = ifelse(ge

Re: [R] Why does R do this?

Dear Nick, The best solution is not to use which() but directy use the logical test. This will work in case the condition is always FALSE and which() returns a integer(0). And it is much faster too. z <- y > 3 y[!z] library(microbenchmark) microbenchmark( y[!y > 3], y[-which(y > 3)] ) Best r

Re: [R] backslash in xtable (generate latex code from R)

You could use kable() from the knitr package. kable(mytable, format = "latex", escape = FALSE) \begin{tabular}{l} \hline $\beta_0$\\ \hline aa\\ \hline bb\\ \hline cc$\alpha_1$\\ \hline \end{tabular} ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INST

Re: [R] Need guidance on summarizing time data

Dear Bruce, I think this should be straightforward with tidyverse. If not please provide a small reproducible data set with dput(). library(tidyverse) count(Active, Time) count(Active, Date, Time) Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Fl

Re: [R] Secondary y axis in ggplot2: did not respond when change its y-axis value

The second y-axis in ggplot2 is only intended to relabel an axis with a fixed transformation. E.g. one axis in degree Celcius and one in Kelvin, km and miles, ... It does not rescale the variables. It looks like you want to display two variables with unrelated units on the same y-axis. That is not

Re: [R] ggplot error of "`data` must be a data frame, or other object coercible by `fortify()`, not an S3 object with class rxlsx"

eth is not a dataframe but of the class rxlsx. You'll need to convert eth into a dataframe. ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INSTITUUT VOOR NATUUR- EN BOSONDERZOEK / RESEARCH INSTITUTE FOR NATURE AND FOREST Team Biometrie & Kwaliteitszorg /

Re: [R] Live Online Training for High School Teachers and Students

Dear Tracy, Maybe a workshop of Data Carpentry (https://datacarpentry.org/) might be relevant. Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INSTITUUT VOOR NATUUR- EN BOSONDERZOEK / RESEARCH INSTITUTE FOR NATURE AND FOREST Team Biometrie

Re: [R] inconsistencies between HLM and clmm results for the signs for the coefficients

Dear Debbie, Have a look at the vignette of the ordinal package. Look for the equation that defines \eta. And note the minus sign associated with \beta. You'll need to find the equation used in HLM and compare it with the equation from ordinal. Best regards, ir. Thierry Onkelinx Statisticus / St

Re: [R] getting to a point where packages are installed and "ready to use", without unnecessarily reinstalling packages

Dear Kelly, Have a look at the renv package (https://CRAN.R-project.org/package=renv). Once setup, your code reduces to renv::restore() Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INSTITUUT VOOR NATUUR- EN BOSONDERZOEK / RESEARCH INSTI

Re: [R] GAM with binary predictors

Dear Sacha, use glm() in this case. I'd rather code the covariable as TRUE / FALSE or as a factor. Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INSTITUUT VOOR NATUUR- EN BOSONDERZOEK / RESEARCH INSTITUTE FOR NATURE AND FOREST Team Biome

Re: [R] Debugging Rmarkdown

Dear Patrick, This is not easy to debug without a reprex I would check the content of zzz and wide.i in the loop str(wide.i) zzz <- rbind(zzz, wide.i) str(zzz) Note that the Rmd always runs in a clean environment. This might explain the difference Best regards, ir. Thierry Onkelinx Statistic

Re: [R] Simulate I x J contingency tables using correlation coefficient using bivariate normal distribution

Dear Sripriya, Step 1. generate random values from a multivariate normal distribution Step 2. convert the random values into probabilities Step 3. convert the probabilities into values from the target distribution library(mvtnorm) n <- 1e3 correl <- 0.9 lambda <- c(10, 50) sigma <- matrix(correl,

Re: [R] linear mixed model required for the U.S. FDA

Dear Helmut, The mixed models list is more suitable for this kind of question. I'm forwarding it to that list. Please send any follow-up to that list instead of the general R help list. If I understand correctly, you'll need a different variance term for both treatments (the within subject for T

[R] mutate() error

Dear all, I'm using this in a function inside a package. The code works fine when I run it directly. df %>% mutate(!!id_column := row_number()) When I run the function I get this error: Error in mutate(., `:=`(!!id_column, row_number())) : argument "mutate" is missing, with no default df is

Re: [R] stats::power.t.test error

Think about this. What is the null hypothesis? What is the alternative? What are their distributions? What is the probability that you get a value from the alternative when the null hypothesis holds and vice versa? Then think again about the relevance of your alternative hypothesis. You'll get a be

Re: [R] [R-sig-ME] lmer error: number of observations <= number of random effects

Dear Srinidhi, You are trying to fit 1 random intercept and 2 random slopes per individual, while you have at most 3 observations per individual. You simply don't have enough data to fit the random slopes. Reduce the random part to (1|ID). Best regards, Thierry ir. Thierry Onkelinx Statisticus

Re: [R] to create a new variable based on values in other variables

Dear Ding, It seems that you are looking for the ifelse() function. Clear use of pmax() and pmin() reduces the number of if statements. m1 <- c(12, 23, 22, 23) m2 <- c(23, 23, 3, 5) Ravg <- ifelse( pmax(m1, m2) == 23, pmin(m1, m2), (m1 + m2) / 2 ) Best regards, ir. Thierry Onkelinx Stati

Re: [R] Package renv and Rscript

Dear Ben, I think that you first need to go to your project and then start Rscript from that location. renv() needs to pick up the .Renviron file located at the root of your project. Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INSTITUU

Re: [R] Rstudio does not run on ubuntu 20

Dear Luigi, This is rather an RStudio problem than an R problem. I suggest contacting RStudio or their community help forum at https://community.rstudio.com/ Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INSTITUUT VOOR NATUUR- EN BOSONDE

Re: [R] unique scale color ggplot2

Dear Catalin, use scale_fill_gradient() and set fixed limits ggplot(df1, aes(x=as.factor(spei), y=as.factor(month), fill = cut(cor, zCuts))) + geom_tile() + scale_fill_gradient(limits = c(-0.7, 0.7)) Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Governmen

Re: [R] unique scale color ggplot2

~ John Tukey /// <https://www.inbo.be> Op di 7 jul. 2020 om 18:42 schreef Rasmus Liland : > On 2020-07-07 12:44 +0200, Thierry Onkelinx via R-help wrote: > > Op di 7 jul. 2020 om 12:02 schreef Catalin Roibu >: > > > &

Re: [R] lme4 - Mixed Model question

Dear Axel, A few quick remarks. 1) This question is more suitable for r-sig-mixed-models. 2) A mixed model needs much more than two groups. If you have only a few groups, consider a glm with group interactions. 3) Setting the value of a covariate to zero for a set of observations has the same eff

Re: [R] Reorganize the data (dplyr or other packages?)

You are looking for tidyr::pivot_longer() Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INSTITUUT VOOR NATUUR- EN BOSONDERZOEK / RESEARCH INSTITUTE FOR NATURE AND FOREST Team Biometrie & Kwaliteitszorg / Team Biometrics & Quality Assuranc

Re: [R] Reorganize the data (dplyr or other packages?)

> melt(x,id.vars="date",measure.vars=c("down","uc","up"),variable.name > ="direction",value.name="percentage") > > HTH, > Eric > > > On Mon, Aug 17, 2020 at 10:49 AM Thierry Onkelinx via R-help < > r-help@r-project.org>