[mlpack-svn] [MLPACK] #209: Armadillo matrix transposition doubles memory usage for a matrix
MLPACK Trac
trac at coffeetalk-1.cc.gatech.edu
Fri Mar 9 20:27:43 EST 2012
#209: Armadillo matrix transposition doubles memory usage for a matrix
------------------------------+---------------------------------------------
Reporter: jcline3 | Owner: rcurtin
Type: defect | Status: new
Priority: major | Milestone:
Component: armadillo sparse | Keywords:
Blocking: | Blocked By:
------------------------------+---------------------------------------------
This was with a large CSV loaded via the kmeans executable, profiled using
valgrind's massif tool.
This isn't a big problem for small matrices, but when it uses several GB
of memory...
One possibility: https://en.wikipedia.org/wiki/In-
place_matrix_transposition
Another might be to modify the way Armadillo loads files?
--
Ticket URL: <https://trac.research.cc.gatech.edu/fastlab/ticket/209>
MLPACK <www.fast-lab.org>
MLPACK is an intuitive, fast, and scalable C++ machine learning library developed by the FASTLAB at Georgia Tech under Dr. Alex Gray.
More information about the mlpack-svn
mailing list