kose-y/SparseKmeansFeatureRanking.jl

Incorrect Variable Usage

Closed this issue · 1 comments

There is mistype in the runtests.jl file on lines 37, 38, 74, 75, 122, and 144 (lines 122 and 144 are currently commented out). In each, there are functions which should take the 'sparsity' variable as an input, but the variable 'm' is inserted instead. The lines of code appear as follows:

Lines 37-38:
@time (classout1, center1, selectedvec1, WSSval1, TSSval1) = ref_sparsekmeans1(X1, class1, classes,m); @time (classout2, center2, selectedvec2, WSSval2, TSSval2) = ref_sparsekmeans2(X2, class2, classes,m);

Lines 74-75:
@time (classout1_, center1_, selectedvec1_, WSSval1_, TSSval1_) = SKFR.sparsekmeans1(X1, m); @time (classout2_, center2_, selectedvec2_, WSSval2_, TSSval2_) = SKFR.sparsekmeans2(X2, m);

Line 122:
@time (classout3,aa,bb,cc,dd)=ref_sparsekpod(copy(y'),classes,m)

Line 144:
@time (classout3_,aa_,bb_,cc_,dd_)=SKFR.sparsekpod(y,classes,m, false, 1)

In each, the variable 'm' should be replaced with 'sparsity' to match the documentation for each function.

Thank you for reporting. I will update the variable names for better readability.