***** test
 x = [1, 2, 3; 4, 5, 6; 7, 8, 9; 3, 2, 1];
 y = ["a"; "a"; "b"; "b"];
 a = fitcknn (x, y, "NSMethod", "exhaustive", "CrossVal", "on");
 assert (class (a), "ClassificationPartitionedModel");
 assert ({a.X, a.Y, a.Trained{1}.NumNeighbors}, {x, y, 1})
 assert (a.ModelParameters.NSMethod, "exhaustive")
 assert (a.ModelParameters.Distance, "euclidean")
 assert ({a.Trained{1}.BucketSize}, {50})
warning: One or more of the unique class values in the stratification variable is not present in one or more folds.
warning: called from
    cvpartition at line 764 column 19
    crossval at line 1701 column 9
    fitcknn at line 354 column 7
    __test__ at line 5 column 2
    test at line 685 column 11
    /tmp/tmp.jyxrHtkgcj at line 782 column 2

!!!!! test failed
ClassificationKNN: not all 'ClassNames' are present in Y.