Added support for Mantid Jacobians and created a Mantid development parser

[AnthonyLim23]

Description of work

This PR adds Jacobians to the mantid parsers. However, it does not work for multistart problems.

It also adds a new mantiddev parser, which removes the advantage from the mantid software when evaluating mantid functions. This does not remove the advantage from mantid jacobians.
Fixes #1228

Testing Instructions

I can send some test problems that use both mantid and mantiddev. It also shows a comparison with nist problems and has similar order of results.

Checklist

• The title is of a format appropriate for a line in future release notes.
• I have added the appropriate tags to the PR.
• My PR represents one logical piece of work.
• My PR fully fixes the issue linked. If new issues have been created, what are they?
• I have added the appropriate tests to cover code that has been added.
• I have updated the documentation in the relevant places to cover the changes.

[github-actions]

Unit & System Test Results

  4 files  ±0    4 suites  ±0   5m 57s ⏱️ +5s
477 tests +7  469 ✅ +7   8 💤 ±0  0 ❌ ±0 
908 runs  +7  853 ✅ +7  55 💤 ±0  0 ❌ ±0 

Results for commit 3c602fa. ± Comparison against base commit fc83cde.

This pull request removes 14 and adds 21 tests. Note that renamed tests count towards both.

fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected0-cutest-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/cutest/start_end_x.SIF]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected0-nist-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/nist/start_end_x.dat]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected1-cutest-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/cutest/basic.SIF]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected1-nist-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/nist/basic.dat]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected10-hogben-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/hogben/start_end_x.txt]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected11-hogben-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/hogben/basic.txt]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected2-nist-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/nist/start_end_x.dat]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected3-nist-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/nist/basic.dat]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected4-mantid-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/mantid/start_end_x.txt]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected5-mantid-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/mantid/basic.txt]
…

fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_factory[mantiddev-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/mantiddev/basic.txt]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_factory[mantiddev-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/mantiddev/start_end_x.txt]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_function_evaluation[/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/mantiddev/function_evaluations.json-mantiddev]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_hessian_evaluation[/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/mantiddev/hessian_evaluations.json-mantiddev]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_jacobian_evaluation[/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/mantiddev/jacobian_evaluations.json-mantiddev]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected0-cutest-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/cutest/basic.SIF]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected0-nist-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/nist/basic.dat]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected1-cutest-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/cutest/start_end_x.SIF]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected1-nist-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/nist/start_end_x.dat]
fitbenchmarking.parsing.tests.test_parsers.TestParsers ‑ test_parsers[expected10-hogben-/__w/fitbenchmarking/fitbenchmarking/fitbenchmarking/parsing/tests/hogben/basic.txt]
…

♻️ This comment has been updated with latest results.

[AndrewLister-STFC]

Looks good. 1 question about the cached jacobians, but I think that's it

[coveralls]

coverage: 90.225% (+0.07%) from 90.157%
when pulling 3c602fa on 1228_mantid
into fc83cde on master.

[AndrewLister-STFC]

All working now apart from that I can't use lru cache with BiComplex... will have to have a think about how best to do this.

[AndrewLister-STFC]

Seems like the caching isn't needed in numdifftools (identical accuracy tables and these runtimes):

vs

[AndrewLister-STFC]

@AnthonyLim23 I think this is ready? Can you check my understanding is right on the docs update I added: https://fitbenchmarking--1231.org.readthedocs.build/en/1231/users/options/jacobian_option.html

[AnthonyLim23]

@AnthonyLim23 I think this is ready? Can you check my understanding is right on the docs update I added: https://fitbenchmarking--1231.org.readthedocs.build/en/1231/users/options/jacobian_option.html

yep. Looks good to me.