I'd like to use numexpr to calculate soft thresholding:
np.sign(x) * np.maximum(abs(x)-l, 0)
in numexpr, one might want to use std::copysign as an optimisation, but that's
another matter.
Original issue reported on code.google.com by pierre.c...@gmail.com on 7 Jun 2012 at 2:48
Original issue reported on code.google.com by
pierre.c...@gmail.com
on 7 Jun 2012 at 2:48