Modified MonteCarloAbsorptionTest.h to add more unit tests, add
integration tests that are based on scenarios with an analytic result
and remove a couple of integration tests where good system test cover
already in place
Add one more test to SpareInstrumentTest to check the inverse square
weighting

Set up factory methods to create the MCAbsorptionStrategy and MCInteractionVolume
objects on the MonteCarloAbsorption algorithm
Create first MonteCarloAbsorption unit test using new mock classes

Merge pull request #28979 from mantidproject/28982_ImproveErrorCalculationOnMonteCarloAbsCorrections

Improve the error calculation in MonteCarloAbsorption

Reset peak HKL if outside tolerance in IndexPeaks

Pick off a couple more unreliable tests

Fix deprecation warning

Detected by address sanitizer

Drops the test time from >100s to ~2-3s and avoids
timeout when running in parallel with other tests
and loading can be even slower.

Finalised mock classes in MCInteractionVolume and created new IMCInteractionVolume
interface class to make mocking more simple
Started work on setting up mock classes in MonteCarloAbsorptionTest

Merge branch 'unit_tests_ununtu_20_04' of https://github.com/mantidproject/mantid into unit_tests_ununtu_20_04

Fix doctests/packaging for Python 3.8

Fix figure options image tab settings for ws with numeric axis

Stitching tab for ORNL SANS

Previously we were shipping a broken setuptools installation
that would not respond to the egg_info command so when
trying to build a dependency the command would fall over.
Machines with all requirements cached would not require
the setup.py step and would succeed

Doctests are not for verifying accuracy but confirming
the api still works expected.

time.clock was removed in Python 3.8 but
this has been missed as the doctests have
not yet been run through Python 3.8

Lazily Generate Python Algorithm Function Signatures and DocStrings

Add fit output status to Indirect Data Analysis fitting tabs

It is always installed with setuptools so setuptools can assume it will always be there.

Signed-off-by: Steven Hahn <hahnse@ornl.gov>

Signed-off-by: Steven Hahn <hahnse@ornl.gov>

28927 ill d7 signed two theta

Signed-off-by: Steven Hahn <hahnse@ornl.gov>

Signed-off-by: Steven Hahn <hahnse@ornl.gov>

Signed-off-by: Steven Hahn <hahnse@ornl.gov>

Signed-off-by: Steven Hahn <hahnse@ornl.gov>

Vectors are stored as Eigen::Vector3d in the detector info which means
that -0.0 is stored as simply 0.0. This alters the long calculation at
the poles and returns 0 instead of -pi. Don't think it matters though
as long not well defined at the pole

Added methods to allow the dependencies of MonteCarloAbsorption to be
overridden during unit testing with Mock classes
Also shared out some of the free functions in SparseInstrument into the
DetectorInfo and InterpolationOption classes where they can be more
easily mocked

Clarify rotations in documentation

Suppress C++17 deprecation warnings on inheriting std::iterator

RE #28809