I'm new to shogun and I've been told that it's efficient with large datasets. I keep reading that Shogun supports LibSVM data format so I thought it'd be easier to switch.
I noticed that shogun needs training data and labels set separately. In LibSVM's file format they are both contained in one data file. How can I load the exact same data file that I created for LibSVM in Shogun (i.e. without separating data and labels)?
checkout the latest develop branch of shogun toolbox from github. it has now native support for reading libsvm file format.
for more details check the examples