Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

problems with logistic regressions #1

Open
ninegua opened this issue Mar 20, 2017 · 4 comments
Open

problems with logistic regressions #1

ninegua opened this issue Mar 20, 2017 · 4 comments
Assignees
@ehsantn

Comments

@ninegua
Copy link

ninegua commented Mar 20, 2017

I've noticed two problems with logistic_regression:

  1. The second dot operation is not matched and turned into parfor.

  2. Result will be wrong if w -= ... is rewritten as w = w - ... in source code.
@ehsantn
Copy link

ehsantn commented Mar 20, 2017

The second dot is essentially a reduction which we don't support yet. Turning it to parfor is easy once we finalize the reduction format.

The second problem is troubling though. I'll take a look.

@ehsantn
Copy link

ehsantn commented Mar 27, 2017

The logistic regression issue is resolved.

@ninegua
Copy link
Author

ninegua commented Apr 4, 2017

We could turn the general matrix multiply into a parfor over a sequential reduction, or with the recent support on reduction, a parfor over a parfor. Thoughts?

@ehsantn
Copy link

ehsantn commented Apr 4, 2017

Currently, the matrix multiply (first dot) is a parfor over a sequential reduction. I don't see any benefit for the parfor over parfor approach for this code since the reduction dimension is small.

ehsantn pushed a commit that referenced this issue May 13, 2018
@anton-malakhov
Merge pull request #1 from stuartarchibald/pr_2748
190a6bd 
Add tests and configuration capacity for SVML implemented by Stuart
ehsantn pushed a commit that referenced this issue Sep 27, 2018
@stuartarchibald
Merge pull request #1 from anton-malakhov/develop
d7db2e2 
Implemented TBB parfor layer, aligned workqueue signatures
DrTodd13 pushed a commit that referenced this issue Mar 22, 2019
@rjenc29
Merge pull request #1 from stuartarchibald/pr_3850_fix1
dbdd51f 
Fix up WRT PR feedback.
DrTodd13 pushed a commit that referenced this issue Sep 17, 2019
@esc
Merge pull request #1 from esc/typed_list_delete_slice
5d20844 
Typed list delete slice
DrTodd13 pushed a commit that referenced this issue Sep 17, 2019
@asmeurer
Merge pull request #1 from stuartarchibald/pr_4345_cont
fa17d7c 
Fix test_function_incompatible_templates test
DrTodd13 pushed a commit that referenced this issue Oct 4, 2019
@Vyacheslav-Smirnov
Merge pull request #1 from numba/master
3d50c71 
Merge changes from origin repo
DrTodd13 pushed a commit that referenced this issue Oct 22, 2020
@kozlov-alexey
Applying review comments #1
d8c0e37
DrTodd13 pushed a commit that referenced this issue Oct 22, 2020
@c200chromebook
Merge pull request #1 from numba/master
8bc0695 
merg
DrTodd13 pushed a commit that referenced this issue Oct 22, 2020
@kozlov-alexey
Applying review comments #1
55e5f79
DrTodd13 pushed a commit that referenced this issue Oct 30, 2020
@John-Ted
Merge pull request #1 from numba/master
e464068 
Sync with original
DrTodd13 pushed a commit that referenced this issue Apr 7, 2021
@braniii
Merge pull request #1 from stuartarchibald/pr_6822_continued
700c0b3 
More constant inference for 6822.
DrTodd13 pushed a commit that referenced this issue Nov 30, 2021
@kozlov-alexey
Applying review comments #1
ae84006
DrTodd13 pushed a commit that referenced this issue May 1, 2023
@sklam
Merge pull request numba#8238 from kc611/adv_idx_1
26bc501 
Advanced Indexing Support #1
DrTodd13 pushed a commit that referenced this issue Aug 22, 2024
@kc611
Merge pull request numba#9662 from kc611/new_ty_sys
2001717 
Type system implementation #1: Added initial implementation for a new type system using redundancies.
DrTodd13 pushed a commit that referenced this issue Aug 22, 2024
@sklam
Merge pull request numba#9686 from kc611/new_ty_sys_review
e980eba 
Type system implementation #1: Added initial implementation for a new type system using redundancies.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@ehsantn ehsantn

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@ninegua @ehsantn