The current frequent pattern and association rule mining algorithms are not implemented for generality not efficiency and may end up consuming unnecessary CPU and memory in certain cases.
As discussed in #893, we could think about a parameter or alternative function to use columns of 0/1's instead of frozensets to speed up the queries.
The current frequent pattern and association rule mining algorithms are not implemented for generality not efficiency and may end up consuming unnecessary CPU and memory in certain cases.
As discussed in #893, we could think about a parameter or alternative function to use columns of 0/1's instead of frozensets to speed up the queries.