Added memory efficient UnsortedDictVectorizer
When loading really large files for SKLL, I found that temporarily
storing a list of dictionaries to pass to DictVectorizer was frequently
using up huge amounts of memory. You can now call `fit_transform` on an
iterable, and not have to waste the temporary space.
Sorting is done in-place, so it doesn't waste memory.
More useful error message for former; return all zeros for latter.
Also changed integer type to np.intc, which matches scipy.sparse
index types more closely.
Fixes#1903.
Replaced all BSD style licenses with "BSD 3 clause"
Not checked yet!
Removed duplicate "3 clause, 3 clause"
Removed trailing period if exists
Fixed some missed licences, still about 50 to do, but those can be automated
Think I got the last of them.
Apparently me and sed have different ideas of regex.
Found a few more
Dan Blanchard