Improve types #101
Improve types #101
Conversation
Codecov Report
@@ Coverage Diff @@
## master #101 +/- ##
=======================================
Coverage 95.69% 95.69%
=======================================
Files 13 13
Lines 488 488
=======================================
Hits 467 467
Misses 21 21
Continue to review full report at Codecov.
|
| sqrtQ[findall(isequal(1), sqrtQ)] = psi[(unknownsH+1):Int(unknownsH + (r/p)*(p*(p + 1)/2))] | ||
|
|
||
| # Obtain full matrices | ||
| H = gram(sqrtH) | ||
| Q = gram(sqrtQ) | ||
|
|
||
| return H, Q | ||
| return T.(H), T.(Q) |
There was a problem hiding this comment.
This is probably affecting performance as we're doing a cast in every estimation step, perhaps there's a better way?
There was a problem hiding this comment.
hmm true, I think it does not make much difference but let us check it
|
We might want to benchmark this PR to see how much these changes have affected performance |
|
Some quick benchmarking showed me that the code has the same performace. We could study how MOI did for benchmarking performance in their testset |
No description provided.