Raymond Hettinger added the comment:
-1 I don't think this is the right way to use heaps. Also, I don't want to
introduce any O(n) operations for fine-grained changes of a single element
(part of the point of having a heap is to make fine-grained changes cheap).
FWIW, it isn't common to change an element and then call heapify. Instead, the
usual approach is either mark an entry as invalid or keep a pending deletion
list or sets.
resolution: -> rejected
stage: -> resolved
status: open -> closed
Python tracker <rep...@bugs.python.org>
Python-bugs-list mailing list