Details
-
Sub-task
-
Status: Resolved
-
Major
-
Resolution: Fixed
-
None
-
None
-
None
Description
In case of sparse matrices, our RDD collect functionality exhibits unnecessary overhead for (1) repeated sparse row allocations (due to missing exploitation of the number of non-zeros which affects the growth rate), and (2) deep copies of sparse rows (even in cases like a single column block, where this is not required).