BiostatR Blog

Ubuntu 24.01  You must first install terra package before install raster package.

The speed is higher using [[n]] then [["x"]]. unname(["x"]) is the slowest.   > Growth        alpha         beta            M            S            K           x0   0.007984356  0.018129643 81.782205061  2.978223538 13.309442413  5.072000000  > system.time({for(i in 1:1E6) K <- Growth[[1]]})    user  system elapsed    0.035   0.000   0.035  > system.time({for(i in 1:1E6) K <- Growth[["alpha"]]})    user  system elapsed    0.083   0.002   0.086  > system.time({for(i in 1:1E6) K <- Growth[1]})    user  system elapsed    0.145   0.004   0.150  > system.time({for(i in 1:1E6) K <- Growth["alpha"]})    user  system elapsed    0.150   0.004   0.154  > system.time({for(i in 1:1E6) K <- unname(Growth[1])})    user  system elapsed    1.122   0.007   1.135  > system.time({for(i in 1:1E6) K <- unname(Growth["alpha"])})    user  system elapsed    1.144   0.011   1.158 

In MacOSX, if you have homebrew installed: To prevent an error during compilation of packages, modify these two files: sudo vi /Library/Frameworks/R.framework/Versions/Current/Resources/etc/Makeconf In "Makeconf" change the line beginning by: FLIBS =  by: FLIBS =  -L/opt/homebrew/Cellar/gcc/14.2.0/lib/gcc/current/gcc/aarch64-apple-darwin23/14 -L/opt/homebrew/Cellar/gcc/14.2.0/lib/gcc/current -lgfortran -lquadmath -lm and the one beginning by: FC = by FC = /opt/homebrew/bin/gfortran -arch arm64 Create a file  sudo vi ~/.R/Makevars with  FC = /opt/homebrew/Cellar/gcc/14.2.0/bin/gfortran F77 = /opt/homebrew/Cellar/gcc/14.2.0/bin/gfortran FLIBS = -L/opt/homebrew/Cellar/gcc/14.2.0/lib/gcc/current The value 14.2.0 must be changed depending on the current version. I try this but it didn't work: FC = $(find /opt/homebrew/Cellar/gcc/* -maxdepth 0 -type d)"/bin/gfortran" F77 = $(find /opt/homebrew/Cellar/gcc/* -maxdepth 0 -type d)"/bin/gfortran" FLIBS = -L$(find

If R crashes and asks to install  additional tools to compile code, try: options(buildtools.check = function(action) TRUE)

Install the last version of fftw to have full possibilities of imager package: Go directly at the end of that post to have the solution ! Check the last version of  fftw here http://www.fftw.org wget http://www.fftw.org/fftw-3.3.10.tar.gz tar -xf fftw-3.3.10.tar.gz cd fftw-3.3.10 ./configure make sudo make install cd .. rm fftw-3.3.10.tar.gz rm -r fftw-3.3.10 then in R install.packages("imager") It fails with imager 1.0.1 with the following error : g++ -std=gnu++11 -shared -L/usr/lib/R/lib -Wl,-Bsymbolic-functions -flto=auto -ffat-lto-objects -flto=auto -Wl,-z,relro -o imager.so RcppExports.o colourspace.o coordinates.o display.o drawing.o filtering.o hough.o imgraphs.o interact.o interpolation.o morphology.o reductions.o transformations.o utils.o wrappers.o -fopenmp -lX11 -Dcimg_use_fftw3 -L/usr/local/lib -lfftw3 -ltiff -L/usr/lib/R/lib -lR /usr/bin/ld: /usr/local/lib/libfftw3.a(assert.o): warning: relocation against `stdout@@GLIBC_2.2.5' in read-only section `.text'

You cannot compare the AIC or BIC when fitting to two different data sets i.e. 𝑌 and 𝑍. You only can compare two models based on AIC or BIC just when fitting to the same data set. Have a look at Model Selection and Multi-model Inference: A Practical Information-theoretic Approach (Burnham and Anderson, 2004). They mentioned my answer on page 81 (section 2.11.3 Transformations of the Response Variable): Investigators should be sure that all hypotheses are modeled using the same response variable (e.g., if the whole set of models were based on log(y), no problem would be created; it is the mixing of response variables that is incorrect). Akaike (1978, pg. 224) describes how the AIC can be adjusted in the presence of a transformed outcome variable to enable model comparison. He states: “the effect of transforming the variable is represented simply by the multiplication of the likelihood by the corresponding Jacobian to the AIC ... for the case of log{𝑦(𝑛)+1}, it is −2 ⋅∑log{𝑦(𝑛)+1},