mikera
Bug in default inverse?
$ lein try net.mikera/core.matrix net.mikera/vectorz-clj
user=> (require '[clojure.core.matrix :as m])
user=> (def a (m/matrix [[0 1 0] [1 1 0] [1 -1 -1]]))
user=> (m/inverse a)
#<NDArrayDouble [[0.0 0.0 0.0] [-0.0 1.0 -0.0] [-0.0 -1.0 -1.0]]>
user=> (m/mmul a (m/inverse a))
[[0.0 1.0 0.0] [0.0 1.0 0.0] [0.0 0.0 1.0]]
user=> (m/set-current-implementation :vectorz)
user=> (def a (m/matrix [[0 1 0] [1 1 0] [1 -1 -1]]))
user=> (m/inverse a)
#<Matrix33 [[-1.0,1.0,0.0],[1.0,-0.0,0.0],[-2.0,1.0,-1.0]]>
user=> (m/mmul a (m/inverse a))
#<Matrix33 [[1.0,0.0,0.0],[0.0,1.0,0.0],[0.0,0.0,1.0]]>
Any progress on this? I also have a case where the inverse matrix becomes singular:
(def a m/matrix [[1 0 0 0.0 0 0.0]
[1 0 1 0.0 0 1.0]
[1 1 0 1.0 0 0.0]
[1 0 1/2 0.0 0N 0.25]
[1 1/2 1/2 0.25 1/4 0.25]
[1 1/2 0 0.25 0N 0.0]])
(det a) ; => 0.015625
(det (inverse a)) ; => IllegalArgumentException lu-decompose can't decompose singular matrix ...
I tested it on both 0.34 and 0.35. Vectorz works fine, as above.
Don't have time to look at this myself right now. I use Vectorz basically all the time.... might just disable this functionality in case it hurts someone.
I'd vote for that option then, I was pretty confused and lost quite a bit of time figuring out what the problem was. It is the default implementation after all.
Hmmm I found a better option I think. Try to use Vectorz if available, throw an error otherwise. Sound sensible?
Yes, even better, unless the transparent conversion introduces a problem elsewhere. The documentation for inverse should mention it too.
Leaving open in case anyone wants to take on fixing / rewriting the default implementation.