You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Reduce bundle size (sparse-array-rled is nontrivial)
More exact asymptotic performance guarantees: O(log(# leaves)) for slightly more leaves, instead of O(log(# leaves) + (max size of a SparseIndices internal rep)).
Cons:
Probably need to merge leaves for reasonable efficiency. Can be lazy (e.g. not between parents) - will always be patched up by reloading anyway.
Might increase memory usage due to more leaves, which means duplicating LeafNode fields. Need to benchmark this.