Rgtsvm icon indicating copy to clipboard operation
Rgtsvm copied to clipboard

The R package for SVM with GPU architecture based on the GTSVM software

Results 12 Rgtsvm issues
Sort by recently updated
recently updated
newest added

My situation: -Ubuntu 18.04 -nvidia-driver-410 (video driver) -GeForce 940M (graphic card) -CUDA 10.1 installed from the *.run When I ran the following command: `R CMD INSTALL --configure-args="--with-cuda-home=$CUDA_HOME --with-boost-home=/usr" Rgtsvm` The...

Got the following error message during installation. Found this issue reported for another package which seemed similar: https://github.com/sirfz/tesserocr/issues/130 The solution stated in that package was to run: sudo apt-get install...

Got Rgtsvm installed and tested it out on a multi class classification task. Training dataset has 12713 samples with 26 variables for each. Rgtsvm is running 1/3rd slower than e1071....

When installing the package I get the following warnings right at the start: ``` > devtools::install_github("NathanSkene/Rgtsvm/Rgtsvm", configure.args="--with-cuda-home=/usr/local/cuda-10.1 --with-boost-home=/home/nskene/R/x86_64-pc-linux-gnu-library/3.6/BH/" ) Downloading GitHub repo NathanSkene/Rgtsvm@master ✔ checking for file ‘/tmp/RtmpeHgXUe/remotesa7b97b696c58/NathanSkene-Rgtsvm-d433558/Rgtsvm/DESCRIPTION’ ... ─...

The way that compilation arguments should be passed to install_github seems to have changed. It now works as: devtools::install_github("NathanSkene/Rgtsvm/Rgtsvm", configure.args="--with-cuda-home=/usr/local/cuda-10.1 --with-boost-home=/usr/local/boost_1_71_0" ) As opposed to how it is stated in...

I'm using RgtSvm to make SVM models for large dataset (nrows=2500000). The problem occurs only with particular number of columns. For example: - 1-10 columns with 2,5 millions of rows...

Hi, Just wondering if there are any plans for a Windows version of the package, and if so, what is the timeline? Thank you, Adam Simkowski

I try to specify sample weights with class.weights, it didn't work. Is there a way to specify sample weights? Or is it not yet implemented?

It should be possible the use of install.packages("Rgtsvm")

I would prefer to use caret as the primary interface for modelling. As Rgtsvm is declared to be compatible with e1071 package, what is the supported way for caret to...