Closed gagolews closed 4 years ago
Computing mst_dist, mst_ind is the slowest part
mst_dist, mst_ind
Getting nn_dist, nn_ind (needed to compute d_core before mst_dist, mst_ind) is also slow (M>1)
nn_dist, nn_ind
d_core
compute_full_tree, postprocess, gini_threshold, n_clusters are only applied after mst_dist, mst_ind are computed
compute_full_tree
postprocess
gini_threshold
n_clusters
Changing X, M, affinity, cast_float32, exact require mst_dist, mst_ind be recomputed
X
M
affinity
cast_float32
exact
However, decreasing M only does not invalidate nn_dist, nn_ind - at least d_core could be generated faster.
Computing
mst_dist, mst_ind
is the slowest partGetting
nn_dist, nn_ind
(needed to computed_core
beforemst_dist, mst_ind
) is also slow (M>1)compute_full_tree
,postprocess
,gini_threshold
,n_clusters
are only applied aftermst_dist, mst_ind
are computedChanging
X
,M
,affinity
,cast_float32
,exact
requiremst_dist, mst_ind
be recomputedHowever, decreasing
M
only does not invalidatenn_dist, nn_ind
- at leastd_core
could be generated faster.