def test_pairwise_distance_numpy_set(): """ Test pairwise distance: using the Numpy (dense matrix) implementation for jaccard set coef """ D = pairwise_proximity(W, metric='jaccard_set') true = np.array([ [ 1., 0.6, 0.4, 0.2, ], [ 0.6, 1., 0.75, 0.5, ], [ 0.4, 0.75, 1., 0.66666667,], [ 0.2, 0.5, 0.66666667, 1., ], ], dtype=float) assert np.isclose(D, true).all()
def test_pairwise_distance_numpy_weighted(): """ Test pairwise distance: using Numpy (dense matrix) using weighted jaccard """ D = pairwise_proximity(W, metric='weighted_jaccard', min_support=10) true = np.array([ [ 1., 0.6, 0.3, 0.1], [ 0.6, 1., 0., 0. ], [ 0.3, 0., 1., 0. ], [ 0.1, 0., 0., 1. ], ], dtype=float) assert np.isclose(D, true).all()
def test_pairwise_distance_numpy_bitwise(): """ Test pairwise distance: using the Numpy (dense matrix) implementation for jaccard bitwise coef """ D = pairwise_proximity(B, metric='jaccard_bitwise') true = np.array([ [ 1., 0.75, 0.5, 0.25 ], [ 0.75, 1., 0.66666667, 0.33333333], [ 0.5, 0.66666667, 1., 0.5 ], [ 0.25, 0.33333333, 0.5, 1. ], ], dtype=float) assert np.isclose(D, true).all()
def test_pairwise_distance_sparse_set(): """ Test pairwise distance: using the Scipy (sparse matrix) implementation for jaccard set coef """ W_sparse = csr_matrix(W) D = pairwise_proximity(W_sparse, metric='jaccard_set') true = np.array([ [ 1., 0.75, 0.5, 0.25 ], [ 0.75, 1., 0.66666667, 0.33333333], [ 0.5, 0.66666667, 1., 0.5 ], [ 0.25, 0.33333333, 0.5, 1. ], ], dtype=float) assert np.isclose(D.todense(), true).all()
def test_pairwise_distance_sparse_weighted(): """ Test pairwise distance: using the Scipy (sparse matrix) implementation for jaccard weighted coef """ W_sparse = csr_matrix(W) D = pairwise_proximity(W_sparse, metric='jaccard_weighted') print D.todense() true = np.array([ [ 1., 0.6, 0.3, 0.1], [ 0.6, 1., 0., 0. ], [ 0.3, 0., 1., 0. ], [ 0.1, 0., 0., 1. ], ], dtype=float) assert np.isclose(D.todense(), true).all()
