Testing how sparse datasets that are only available as RData file could be converted to the mldata.org hdf5 format. This is not working as the hdf5 specification stated on mldata.org is not recognized from the there parser.

convert_mm_file_to_mldata_h5.py

from scipy.io.mmio import mminfo,mmread,mmwrite
import numpy as np
import h5py
import os

os.chdir('/home/mane/workspace/benchmark/')

print mminfo('InternetAd')

X = mmread('InternetAd.mtx')

y = np.loadtxt('InternetAd.target', skiprows=0)

f = h5py.File('InternetAd.h5', 'w')

# HDF5 attributes on the root level
f.attrs['mldata'] = 0
f.attrs['name'] = 'InternetAd'
f.attrs['comment'] = 'Internet-Ad [Kushmerick, 1999]: document classification problem with \
mostly binary features.'


#HDF5 group 'data_descr'
f.create_group('/data_descr')
f['/data_descr/ordering'] = np.array(['label', 'data'])

#HDF5 group 'data'
f.create_group('/data')
f['/data/data'] = X.data
f['/data/data_indices'] =  X.tocsc().indices
f['/data/data_indptr'] = X.tocsc().indptr
f['/data/label'] = y

f.close()

convert_R_sparse_data_to_mm.R

load('InternetAd.RData')

file = file.path(getwd(), 'InternetAd.mtx')

library(Matrix)
writeMM(InternetAd$x, file=file)

file = file.path(getwd(), 'InternetAd.target')

write(InternetAd$y, file=file, ncolumns=1)



load('NewsGroup.RData')

file = file.path(getwd(), 'NewGroup.mtx')

library(Matrix)
writeMM(NewsGroup$x, file=file)

file = file.path(getwd(), 'NewsGroup.target')

write(NewsGroup$y, file=file, ncolumns=1)