UCL / UCL/STIR

Objective functions should not compute the sensitivities upon `set_up`

Open
#912 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
C++
Stars
160
Forks
113
Avg merge
12d 15h
Merged PRs (30d)
1

Description

Utilities such as OSSPS,compute_sqrt_Hessian_row_sum, demo4 and demo5 utilise Poisson objective functions but do not need the system matrix sensitivities (A^T 1), yet they are computed. This additional and unnessicery computation slows down these utilities.

Perhaps with the changes being made in #893, the sensitivity do not need to be computed during set_up. Prior to this PR, the sensitivities were used in the subset-gradient computation, grad = A_m^T [ y_m/ (A_m(x) + b_m) ] - sensitivity_m making this suggestion infeasable.

When #893 is merged, it could be that the sensitivities are only computed if objective_function.get_subset_sensitivity(subset_num); is called.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reviewing the objective-function setup and the changes proposed in issue #893. Check the OSSPS, compute_sqrt_Hessian_row_sum, demo4, and demo5 call paths, then verify that sensitivities are computed only when get_subset_sensitivity(subset_num) is called and that subset-gradient behavior remains correct.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
performance
Issue type
Refactor
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.