when calling a SVD with FloatSingularValueDecompositionDC on a SparseFloatMatrix2D (here m1)
FloatSingularValueDecompositionDC usv = new FloatSingularValueDecompositionDC(m1, true, false)
one gets the following error:
Exception in thread "main" java.lang.ClassCastException: cern.colt.map.tfloat.OpenLongFloatHashMap cannot be cast to [F
at cern.colt.matrix.tfloat.algo.decomposition.FloatSingularValueDecompositionDC.<init>(Unknown Source)
if one uses a DenseFloatMatrix2D for m1 instead, it works.
it works as well with SparseFloatMatrix2D and FloatSingularValueDecomposition, but takes much more time.