fitbenchmarking / fitbenchmarking / 31392900384
94%

DEFAULT BRANCH: master
Ran
Jobs 2
Files 82
Run time 1min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 93.516% (+0.03%) from 93.488%
31392900384

push

github

web-flow 
Added multifit capability (fitting multiple datasets) to more minimizers other than Mantid (#1637)

* Initial work

* Modify final_params to allow multifit

* Add comment on data_x

* Update functions in base_controller

* Fix issue with data_x

* Comment out function for now

* Ensure params is a list of lists

* Modify eval_r and jac_res in weighted_nlls_cost_func (unsure whether this should be done somewhere else)

* Add ties, though not being used right now

* Move for loop to eval_model

* latest multifit progress (still WIP)

* Resetting par_names and param_names at every iter of the minimizers for loop

* Trying to get eval_r to run when called by eval_chisq

* reset initial_params in multifit_cleanup

* Remove unnecessary print statements and comments

* Remove more unneeded print statements

* Add some useful comments

* Fix issue with params in eval_model

* Fix issue with residuals plots

* Remove print statements

* Set set.multifit to False in fitting_problem

* Remove logger.info to avoid getting "utf-8 codec can't decode byte 0xe2 .."

* remove all print statements

* Add a docstring for the new eval_r

* Linting fixes

* Add tests for multifit_init and multifit_cleanup

* potential fix to mantid tests

* more tests for multifit work

* Potential fix to failing tests due to conflicts with mantid multifit

* potential fix again to more mantid tests

* Fix test by changing eval_r to eval_r_single_dataset

* remove mantid_ties and get_ties_str function

* Update regression tests and add workaround for editing problem._param_names

* fix failing test

* update regression tests

* Add capability to handle bounds

* pin sasmodels version to the latest working

* Add docs

* Add validate_multifit in base_controller

* Add tests for validate_multifit

* Update docs to specify incompatible options for multifit

* Make sure multifit_cleanup runs also when fit fails

* remove tracker changes, to be dealt with in another PR

* remove redundant comm... (continued)

136 of 160 new or added lines in 19 files covered. (85.0%)

7 existing lines in 5 files now uncovered.

5423 of 5799 relevant lines covered (93.52%)

1.65 hits per line

Uncovered Changes

Lines Coverage ∆ File
5
0.0
 0.0% fitbenchmarking/controllers/gradient_free_controller.py
3
17.5
 -0.22% fitbenchmarking/controllers/bumps_controller.py
3
0.0
 0.0% fitbenchmarking/controllers/ceres_controller.py
3
97.68
 -0.96% fitbenchmarking/core/fitting_benchmarking.py
3
90.91
 -4.33% fitbenchmarking/cost_func/nlls_base_cost_func.py
2
0.0
 0.0% fitbenchmarking/controllers/galahad_controller.py
2
0.0
 0.0% fitbenchmarking/controllers/lmfit_controller.py
2
0.0
 0.0% fitbenchmarking/controllers/theseus_controller.py
1
0.0
 0.0% fitbenchmarking/controllers/gofit_controller.py

Coverage Regressions

Lines Coverage ∆ File
2
0.0
 0.0% fitbenchmarking/controllers/ceres_controller.py
2
0.0
 0.0% fitbenchmarking/controllers/gradient_free_controller.py
1
17.5
 -0.22% fitbenchmarking/controllers/bumps_controller.py
1
0.0
 0.0% fitbenchmarking/controllers/galahad_controller.py
1
0.0
 0.0% fitbenchmarking/controllers/gofit_controller.py
Jobs
ID Job ID Ran Files Coverage
1 31392900384.1 82
75.82
 GitHub Action Run
2 31392900384.2 82
89.39
 GitHub Action Run
Source Files on build 31392900384