Move ISIS indirect reduction into Mantid and write proper system tests for it

[Alex Buts]

Mantid currently uses DirectInelastic reduction algorithms which is used for converting data to energy transfer. The integrity of these is verified by the system tests.

These algorithms are the basis for customized reduction, actually used in ISIS excitation group to process the experimental results.

The custom algorithms mainly rely on Mantid algorithm but override different settings from different places and have some parts of the algorithm implemented differently.

1) The custom algorithm are not covered by system tests and can be easily broken by the changes in Mantid core

2) Two (actually three) sets of default parameters plus possibility to set up parameters from the script itself make it difficult to understand what parameter is actually used.

The purpose of this ticket is to bring all ISIS inelastic reduction script into Mantid and verify its integrity by proper system tests.

Isis reduction script should be very thin layer over verified Mantid reduction scripts.

[Alex Buts]

refs #6667 Instrument parameters modified to take some properties

from dgreduce

Changeset: 7054b8b689062ba1e9a63ded7b93f83c01eb541d

[Alex Buts]

refs #6667 first steps in taking all parameters from dgreduce

and moving them into IDF_Parameters.xml

Changeset: 1d3ead08640467109bf7d2dfd8b0763ba92a6743

[Alex Buts]

refs #6667 Changed ExperimentInfo::populateInstrumentParameters

to understand booleans, though no unit tests found for populateInstrumentParameters to check this stuff works properly

Changeset: a3ad11252080a8c09bfc31d54a756d4dcc071b31

[Alex Buts]

refs #6667 Exported parameter type from the parameter map

Very ugly done, as python should be able to retrieve the value and type together. But can not find better way at the moment.

Changeset: 6fa81f435d321d7321ad49a03eb8d4581cfb3f67

[Alex Buts]

refs #6667 Minor changes to export parameter type

and substitution dictionary in DirectEnergyConversion

Changeset: 7cc4636d2a76931ca3b300c525e30d826d29589e

[Alex Buts]

refs #6667 Failed attempt to build getParameterType test

Changeset: be8737adb1cb6e8d4dcb3bead11faeaef4cbbe2a

[Alex Buts]

refs #6667 finally correct test verifying the component type name

Changeset: ea809f14f70cfdb57f9513f58346ce107aaef727

[Alex Buts]

refs #6667 Hopefully working solution with Parameter-based file

where all default keys are obtained from Par.xml. + int parameter is added to xml parcer.

Need tests and routines, accepting parameters from the arguments.

Changeset: fd1045910f9e6cce0b5b025440581f1b42d8ebfb

[Alex Buts]

refs #6667 more changes to parameters and some unit tests attempts

Changeset: 22d5bcf8e8e26b6cdbd92235b972dc53afc30052

[Alex Buts]

refs #6667 DirectEnergyConversion accepts external parameters

and have number of unit tests enabled testing this functionality. DAE name property imported from Mantid Facility (for qtiGenie to work)

Changeset: 8e016fb1c6c43027470d44a26f26082f7d9bc5e2

[Alex Buts]

refs #6667 intermediate checkout

some unit tests run and ready to run system tests

Changeset: 568b5c837ac8767a259ef23248aa35971802f503

[Alex Buts]

refs #6667 diagnose seems work

Changeset: 3af63f712d6fb3986bb3aeef2f9203d188f64fcd

[Alex Buts]

refs #6667 Intermediate checkout

Old system test for MAPS runs with modified DirectEnergyConversion

Changeset: 4fe78d10f0085fdf597acfe9ef0a830d9b6af779

[Alex Buts]

refs #6667 modifying system tests to work with new dgreduce

in ISIS

Changeset: 6553bd5dafdd41ad095d090c54699aedb8396f6b

[Alex Buts]

refs #6667 Maps system tests run but do not pass

Both old and new tests failing, which is probably due to different defaults

Changeset: c0e66a5e324bac4801fc10da207f3ff7b035a8aa

[Alex Buts]

refs #6667 Added dgreducer based system test for MAPS

Changeset: a57029385e6dd3a676d8b642e8b043a1ef3b0ea6

[Alex Buts]

refs #6667 All generally works but constants are missing.

Changeset: c6a8e0e7416b9ed452f40a167fc3e0a99ef10835

[Alex Buts]

refs #6667 The run data look very close to each other (MAPS)

but not yet identical -- diag seems a bit different

Changeset: 99a8f64f858ef44bf984d95acdd78af3bfb2e74b

[Alex Buts]

refs #6667 minor changes

mainly related to ongoing work on other branches

Changeset: b20d8cc966ec528dd550cc2db74e2fae013b6815

[Alex Buts]

refs #6667 Maps reduction works

But there is subtle difference in reduction results

the difference does not have big physical sense, as the relative values of differences are small, but still unclear why it is present at all.

The CheckWorkspacesMatch got possibility to analyze relative error together with the absolute error to understand the difference better LoadMask.cpp in the ISIS part has been changed to report loaded masks on debug level rather then notice as the messages in the log complicate understanding of what is going on.

Changeset: d4cbed63b7fc690844d623c9a6ecb920e3983c88

[Alex Buts]

refs #6667 Minor changes for niceness

Changeset: a3a3ba23f03582c28abbb04e30064882c8e61783

[Alex Buts]

refs #6667 Modified parameters in accordance with MAPS

mean real data which MAPS uses -- not the one it has in MAPS parameters file

Changeset: f0e7a7f96c4ed4b10a1878de0af0d82bc47d4eef

[Alex Buts]

refs #6667 minor changes

Changeset: 58e44c54245486abd8b81722e48b22a0a0b5a9f0

[Alex Buts]

refs #6667 Intermediate commit -- debug statements

Changeset: bd79a0f926ec6d56aa056e57e866e5cd3534b64e

[Alex Buts]

refs #6667 Small improvements and attempt to use hardmask only

unsuccessful attempts

Changeset: b8394503dda2351f22edb473afdf6be865a2ef4d

[Alex Buts]

refs #6667 minor changes

Changeset: ae375bed30bccf83ce6f328c145fb3c38da9e43e

[Alex Buts]

refs #6667 Modified Save procedure in DirectEnergyConversion

to use different save routines as lambda functions

plus some small debugging statements which will probably be reverted

Changeset: ee8b3ada5f305d9a70415f75835c823dfa269ce1

[Alex Buts]

refs #6667 Modified getNeutronAtom without exceptions to be really

without exceptions

As it is impossible to start Mantid in Debug mode as it throws thousand exceptions which are then catched, analyzed, reported and one has to wait for 1 min to start trivial program

Changeset: dc6ef632982584d0254bae99d316da082a671f26

[Alex Buts]

refs #6667 Presumably created hardMaskOnly mode in diagnose

removed couple of small test clauses.

Changeset: d981f12f335e45371eae00ab4d4a3c128602f718

[Alex Buts]

refs #6667 Better defined overloaded properties

Changeset: 4c28e762e188775b9bae95b73c0b7d331617425f

[Alex Buts]

refs #6667 Checkpoint.

Trying to reduce monovanadium and it does not work.

Changeset: 96416491a7cf259b4b4035f10a4cd66c341fa02a

[Alex Buts]

refs #6667 Hard Mask only works

Changeset: 197697037a468856345391d3cc8fa37a36e8647b

[Alex Buts]

refs #6667 Implemented default relative mono-vanadium integration

range. There are doubts on if it is correct implementation as seems it was used for different energy ranges. But let's do something and see.

Changeset: c2ba791dd1509323c6b92d18fae8b850fb5f4703

[Alex Buts]

refs #6667 Attempted to implement masks reuse.

Which does not work because of the way xml mask interacts with workspace. left for the time being.

Changeset: e8972f3a51c44f92f175748debb42843ed297569

[Alex Buts]

refs #6667 Intensities of old and new dgreduce coincide in arb units

Changeset: 5c2c5acf182527ef874aeea5493a485ab618f55d

[Alex Buts]

refs #6667 Some comments

Changeset: dc00e5a9f099854567ce64f48b164cc19a66f613

[Alex Buts]

refs #6667 Minor comments and enhancements

Changeset: f081441e33266ba9952fa04455d30eb5cb3b3e88

[Alex Buts]

refs #6667 Added option to request comparing all data to

CheckWorkspacesMatch

Changeset: 0168f2ef7f9c8cdce8c9aca8e63afcf72db076b9

[Alex Buts]

refs #6667 Comments and code niceness.

Changeset: d8342668c12453ea87c5f5d864696faed44577d8

[Alex Buts]

refs #6667 MAPS system test modified to work hopefully right

and produce correct WB cross-section if set up this way (currently it is not and runs against some run data)

Changeset: 6d2de5721e5f066160971a452a5d48241a7d8c5a

[Alex Buts]

refs #6667 LET parameters file has been created and tried with LET

reduction

Changeset: 6cd4f86f82d9e2a47d1d06011f11788d18f8eb78

[Alex Buts]

refs #6667 checkpoint Changed set_instrument_name property

working together with calibration file workspace

Changeset: 7a4012635c12279d8f478e4592b9328984460bb7

[Alex Buts]

refs #6667 Minor checkpoint

Changeset: fa9cc2ee8fcc7f3a87e6e98e4102343647874650

[Alex Buts]

refs #6667 Small changes allowing to run DirectEnergyConversion for

LET

Changeset: 10064f82ad0eea0329453dc37dc1fc63891d54b1

[Alex Buts]

refs #6667 Added provisional parameters for LET MERLIN and MARI

the old remaining as back-up

Changeset: a9ff4638ae5b7e4b8a3d58e441a86cb433bd16ba

[Alex Buts]

refs #6667 Something like working for MARI

but further investigations pending if it is exactly right

Changeset: 2aaf1e64ffcfc46ddfae7f2c93692443b302036b

[Alex Buts]

refs #6667 Modified MAPS and MARI System tests pass

one more workflow test has been added for MARI

Changeset: 82196cf97d5770f90c2d4b16b599995495eb5b6c

[Alex Buts]

refs #6667 Changes to MARI reduction to produce previous results

MARI generates the results which are similar to the one, produced by old reducer. Reasons for difference are generally clear.

Changeset: 7ce4293e1934fa9278d842e128643ace1500f78c

[Alex Buts]

refs #6667 MARI and MAPS system tests pass

modified system tests, compared against vanadium

Changeset: c177bb10e0cc7d3cb07ebae64b18c682a1662f28

[Alex Buts]

refs #6667 Extracted unit tests for the interfaces

and added them to CMakeLists.txt

Changeset: 5513e3000e67fddeba82a35c759d735e2e76c581

[Alex Buts]

refs #6667 System tests modified to work for dgreduce.

Changeset: 6d09eeac5ef01b4a3d7a38990d9492e6da350adb

[Alex Buts]

refs #6667 Correct test if workspace provided to reducer

Changeset: 9b4ab8bdd9b8e662b2cf554d2b164b703dc78d2a

[Alex Buts]

refs #6667 Modified getNeutronAtom without exceptions to be really

without exceptions

As it is impossible to start Mantid in Debug mode as it throws thousand exceptions which are then catched, analyzed, reported and one has to wait for 1 min to start trivial program

Changeset: 0df38d27c4d410c0320cb96774a08ca3818b16d0

[Alex Buts]

refs #6667 Presumably created hardMaskOnly mode in diagnose

removed couple of small test clauses.

Changeset: 596aa803ab64cdbe20b274c861f0511efe3c2d11

[Alex Buts]

refs #6667 Better defined overloaded properties

Changeset: 3711e07c27c001021b4df979b76cbc57e9fb28ab

[Alex Buts]

refs #6667 Checkpoint.

Trying to reduce monovanadium and it does not work.

Changeset: ad59da8b56ed48fe39cb01d760ba1048c2e5170b

[Alex Buts]

refs #6667 Hard Mask only works

Changeset: 2ac4bc0c9d3395edd2ed374d21a9cbaaf1fac280

[Alex Buts]

refs #6667 Implemented default relative mono-vanadium integration

range. There are doubts on if it is correct implementation as seems it was used for different energy ranges. But let's do something and see.

Changeset: 0aa884c9d76a7b9be3ac8bf0edf310eb0bcf318a

[Alex Buts]

refs #6667 Attempted to implement masks reuse.

Which does not work because of the way xml mask interacts with workspace. left for the time being.

Changeset: 5dc8cbb42a0666239d367b34552a7e5dc850ccfa

[Alex Buts]

refs #6667 Intensities of old and new dgreduce coincide in arb units

Changeset: 48d23a944b71168accf16470bbf2b24753f80c57

[Alex Buts]

refs #6667 Some comments

Changeset: 0279359f44107cd3f66ded6b38f379a27f2e0fdd

[Alex Buts]

refs #6667 Minor comments and enhancements

Changeset: 8fd0ed42de26ea9b93f3416abc243590b4203cb6

[Alex Buts]

refs #6667 Added option to request comparing all data to

CheckWorkspacesMatch

Changeset: fb1964e487a39ecddf3ea83cc4b8c2754098aaee

[Alex Buts]

refs #6667 Comments and code niceness.

Changeset: ee45d2074357cab6ba9a9816a90445944903d966

[Alex Buts]

refs #6667 LET parameters file has been created and tried with LET

reduction

Changeset: 19aa4f9d4dd8f1eeb779fbb9b12404800a085206

[Alex Buts]

refs #6667 checkpoint Changed set_instrument_name property

working together with calibration file workspace

Changeset: cb45ffeda2308a46bbca75ab1ff4fb7bbed8cb31

[Alex Buts]

refs #6667 Minor checkpoint

Changeset: eb5063c9234c143a4c0e6bd5a296c162ba48e21a

[Alex Buts]

refs #6667 Small changes allowing to run DirectEnergyConversion for

LET

Changeset: 34a2a5c15118b143d874c59f8bad534df79a2193

[Alex Buts]

refs #6667 Added provisional parameters for LET MERLIN and MARI

the old remaining as back-up

Changeset: 4359af7f11cfa632bc15b44ba643191da50c5089

[Alex Buts]

refs #6667 Something like working for MARI

but further investigations pending if it is exactly right

Changeset: a80b20b51594f6aa7c0a4d075a95da74fdcc3a46

[Alex Buts]

refs #6667 Changes to MARI reduction to produce previous results

MARI generates the results which are similar to the one, produced by old reducer. Reasons for difference are generally clear.

Changeset: 884e2ba501aec758fc7fd1d9e392526f79910dff

[Alex Buts]

refs #6667 MARI and MAPS system tests pass

modified system tests, compared against vanadium

Changeset: 1933b51647408634ceb45ce875e505a05e85cacc

[Alex Buts]

refs #6667 Extracted unit tests for the interfaces

and added them to CMakeLists.txt

Changeset: 49544007feaaac826c11665dcfbab4ed0d395d34

[Alex Buts]

refs #6667 Correct test if workspace provided to reducer

Changeset: 032aae03cf3c2c25f91776981fcdd62a56e67c49

[Alex Buts]

refs #6667 Modified getNeutronAtom without exceptions to be really

without exceptions

As it is impossible to start Mantid in Debug mode as it throws thousand exceptions which are then catched, analyzed, reported and one has to wait for 1 min to start trivial program

Changeset: dc6ef632982584d0254bae99d316da082a671f26

[Alex Buts]

refs #6667 Presumably created hardMaskOnly mode in diagnose

removed couple of small test clauses.

Changeset: d981f12f335e45371eae00ab4d4a3c128602f718

[Alex Buts]

refs #6667 Better defined overloaded properties

Changeset: 4c28e762e188775b9bae95b73c0b7d331617425f

[Alex Buts]

refs #6667 Checkpoint.

Trying to reduce monovanadium and it does not work.

Changeset: 96416491a7cf259b4b4035f10a4cd66c341fa02a

[Alex Buts]

refs #6667 Hard Mask only works

Changeset: 197697037a468856345391d3cc8fa37a36e8647b

[Alex Buts]

refs #6667 Implemented default relative mono-vanadium integration

range. There are doubts on if it is correct implementation as seems it was used for different energy ranges. But let's do something and see.

Changeset: c2ba791dd1509323c6b92d18fae8b850fb5f4703

[Alex Buts]

refs #6667 Attempted to implement masks reuse.

Which does not work because of the way xml mask interacts with workspace. left for the time being.

Changeset: e8972f3a51c44f92f175748debb42843ed297569

[Alex Buts]

refs #6667 Intensities of old and new dgreduce coincide in arb units

Changeset: 5c2c5acf182527ef874aeea5493a485ab618f55d

[Alex Buts]

refs #6667 Some comments

Changeset: dc00e5a9f099854567ce64f48b164cc19a66f613

[Alex Buts]

refs #6667 Minor comments and enhancements

Changeset: f081441e33266ba9952fa04455d30eb5cb3b3e88

[Alex Buts]

refs #6667 Added option to request comparing all data to

CheckWorkspacesMatch

Changeset: 0168f2ef7f9c8cdce8c9aca8e63afcf72db076b9

[Alex Buts]

refs #6667 Comments and code niceness.

Changeset: d8342668c12453ea87c5f5d864696faed44577d8

[Alex Buts]

refs #6667 LET parameters file has been created and tried with LET

reduction

Changeset: 6cd4f86f82d9e2a47d1d06011f11788d18f8eb78

[Alex Buts]

refs #6667 checkpoint Changed set_instrument_name property

working together with calibration file workspace

Changeset: 7a4012635c12279d8f478e4592b9328984460bb7

[Alex Buts]

refs #6667 Minor checkpoint

Changeset: fa9cc2ee8fcc7f3a87e6e98e4102343647874650

[Alex Buts]

refs #6667 Small changes allowing to run DirectEnergyConversion for

LET

Changeset: 10064f82ad0eea0329453dc37dc1fc63891d54b1

[Alex Buts]

refs #6667 Added provisional parameters for LET MERLIN and MARI

the old remaining as back-up

Changeset: a9ff4638ae5b7e4b8a3d58e441a86cb433bd16ba

[Alex Buts]

refs #6667 Something like working for MARI

but further investigations pending if it is exactly right

Changeset: 2aaf1e64ffcfc46ddfae7f2c93692443b302036b

[Alex Buts]

refs #6667 Changes to MARI reduction to produce previous results

MARI generates the results which are similar to the one, produced by old reducer. Reasons for difference are generally clear.

Changeset: 7ce4293e1934fa9278d842e128643ace1500f78c

[Alex Buts]

refs #6667 MARI and MAPS system tests pass

modified system tests, compared against vanadium

Changeset: c177bb10e0cc7d3cb07ebae64b18c682a1662f28

[Alex Buts]

refs #6667 Extracted unit tests for the interfaces

and added them to CMakeLists.txt

Changeset: 5513e3000e67fddeba82a35c759d735e2e76c581

[Alex Buts]

refs #6667 Correct test if workspace provided to reducer

Changeset: 9b4ab8bdd9b8e662b2cf554d2b164b703dc78d2a

[Alex Buts]

refs #6667 Added missing imports to unit tests

Changeset: 77ec783d59f37ba144ec5099496d340fad0f6d50

[Alex Buts]

refs #6667 Fixing previous error

and trying to guess correct CMakeList for python tests

Changeset: cc70258be3534f8685b521a25b3f54b2de0d1e04

[Alex Buts]

refs #6667 temporary disabling failing test

Changeset: 92be120d826aaceb78671f723819fd89df2387dd

[Alex Buts]

refs #6667 Trying to add python library to test to python search path

Changeset: 6257183e3d28c39df8015d53bacc6153ffbaa085

[Alex Buts]

refs #6667 Another attempt to set search path for python test

Changeset: da8b25469e97f434ca99bf58f21f9b4fc2bb7eb2

[Alex Buts]

refs #6667 Playing with python search path in CMakeLists.txt

No luck

Changeset: dbeb44a269a58f73876ebfafe34ab08275365454

[Alex Buts]

refs #6667 last attempt to give proper path to test modules

Changeset: 80903791551cf4a0b546a6d229159dbd5f2c7b39

[Alex Buts]

refs #6667 All back, does not work

Changeset: 055920832d8ee320a78c904968b6901cbb52088d

[Alex Buts]

refs #6667 Added missing file needed for MAPS reduction

Changeset: 11ed88d8f68145863b04070a9084a7a2c91c4930

[Alex Buts]

refs #6667 Fixed DirectInelasticDiagnostic and LoadLotsOfFiles

hopefully

Changeset: 96fb4dc5943268a8b7489454bfa1cd622b032348

[Alex Buts]

refs #6667 Modified WFA helper to understand bool property

this should fix DGSReduction algorithm

Changeset: 2d4a9344e112b573c80ae7c1af56f92d79a9432b

[Alex Buts]

refs #6667 Should fix mysterious error in MARI sum test

Changeset: 151c11e017ee3b31777e6e05cdbe9bd88f3a3019

[Alex Buts]

refs #6667 Some key-words corresponding to Parameters file

Changeset: 763fe727b31e58027dfc48e16e6262c472239615

[Alex Buts]

refs #6667 guessed last parameters providing difference

in DirectInelasticDiagnostic2

Changeset: d60da64f7f8e132db9ea304044e51ce67cc574d8

[Alex Buts]

refs #6667 Minor changes

use correct procedure to find non-run type file path

Changeset: d02604fcbbc45ffc37bc9991ff15797d295cb72b

[Alex Buts]

refs #6667 works correctly with necessary file as run number

also tries to modify SNS reduction to understand new properties.

Changeset: ce9456e354ead12ca0c23cf86b953553909e0fbf

[Alex Buts]

refs #6667 fixed dgreduce test run separately

Changeset: 9bcd9527ea2c3a60b0157029794fa2cea0f2aafe

[Alex Buts]

refs #6667 The change should fix python unit tests

hopefully. And the tests were enabled.

Changeset: 9457c592f32ea8a41f3806dd24dc2bfa18032901

[Alex Buts]

refs #6667 Made unit test comparison almost equal for floating point

As the test fails on Unix otherwise.

Changeset: a01b0cb1f8ae31f09e63bdf81945cda12510c23d

[Alex Buts]

refs #6667 Apparently more places where almost equal does not work on

Unix

Changeset: bb183539afd3139857ceb4ea634e56c42c3e4a79

[Alex Buts]

refs #6667 Added new parameter types to API-1

for old DGS reduction to work

Changeset: 19705bdb3a7d48c59b6fe182b46b8087eafc6803

[Alex Buts]

The ticket branch exists and need merging both in Mantid and SystemTests repositories.

The main task of this ticket is hardly testable.

I took the “trusted” reduction scripts used on ISIS inelastic instruments, extracted the parameters, embedded into these scripts into InstrumentName_Parameters.xml files and put new xml files and remaining shell of the scripts into Mantid. The scripts are built around old DirectEnergyConversion (dgreduce) Mantid script, so I have changed the interface of this script to understand parameters provided in the new way.

Then I’ve verified if the modified Mantid reduction produces almost exactly the same results and the trusted scripts and where it produces different results (MARI the example) me and the instrument scientist understand the reason for the difference and agree about the changes.

System tests run in Mantid were modified to understand new parameters and produce the results on the basis of these parameters. The DirectEnergyConversion scripts were slightly modified to fix some errors but full modification was not event tired from point of view of switching to DGSReduction workflow, developed in SNS.

As all this works, produces reasonable results have been extensively used during last tree weeks, tested and agreed with ISIS instrument scientists, a tester can accept it without big worries.

There are two areas related to the changes in C++/python code where careful tester can try to verify things: 1) The instrument parameter file can now have Boolean and Int parameters in addition to old double and string parameters. The example of changes is in the ticket’s xml files and unit tests are in Python, in DirectEnergyConversionTest.py file. Tester can look on how this code works and if anything can be broken/ improved there. 2) While testing the changes I have added new parameter to CheckWorkspaceMatch algorithm, which allows testing difference in workspaces as relative error. The changes are trivial and I used modified algorithm extensively, but tester can think about ways of verifying this.

[Martyn Gigg]

There are failing system tests that need to be looked at first.

[Alex Buts]

refs #6667 Should fix issue when arb units requested files

necessary for abs units only

Changeset: b8cecf234d2187c1da936c7530386a78fe7a537d

[Alex Buts]

refs #6667 Should fix the issue when abs units files are requested

for arb units run

Changeset: 6d3e1ed171464711d46de21466e1e036a558c232

[Martyn Gigg]

There are a few things with the changes to CheckWorkspacesMatch algorithm that we need to change slightly. Adding the new properties will change the order that they are called in Python so I think we need to move:

• declareProperty("ToleranceRelerr",false, "Treat tolerance as relative error rather then the absolute error");
• declareProperty("CheckAllData",false, "Usually checking data ends wnen first mismatch occurs. This will force algorithm to check all data and print mismatch to debug log");

to the end after

• declareProperty("Result","",Direction::Output);

That way the order will stay the same in Python.

Could you also change the name ToleranceRelerr to ToleranceRelErr, so it is more consistent with normal property style. While you are there could you fix the spelling of wnen->when.

Be aware that ticket #7403 removed the "supported.facilties" key from the properties file. I've merged master to the branch feature/6667_ISISReductionInMantid and updated the code in DirectEnergyConversion to use the new features. Make sure you do

• git checkout feature/6667_ISISReductionInMantid
• git pull origin feature/6667_ISISReductionInMantid

before continuing with any work

[Martyn Gigg]

Add export of Facility* to python.

Needed for using those returned from config.getFacilities. Refs #6667

Changeset: dfc89c4153951e50bd8360b06dca726de748c174

[Alex Buts]

refs #6667 Minor modifications to CheckWorkspacesMatch

to address tester's comments.

Changeset: 39e22fdbc0a5ff1c7a2790312d3077b50e42de83

[Alex Buts]

refs #6667 enabled check_background property

It is apparently used by new SNS reduction, in the meaning similar to background property. It is easy to modify DirectEnergyConversion to enable it for current isis reduction to simplify move

Changeset: 0c3a72f8c2817e3bff73dc884b24f838de26de2c

[Alex Buts]

refs #6667 Fixed unit test affected by check_background changed

Changeset: 9741bfb341668b93202015a94bd2b60c4eb330d2

[Martyn Gigg]

The component stuff is not ideal but the parameter extraction is in a bit of a mess anyway and this hasn't really made it much worse.

The changes to CheckWorkspacesMatch & Component/ParameterMap should really have been done in another ticket that could have been tested in isolation but they look okay.

The systemtests are passing and the new reference files seem to make sense (visually) anyway and are required because of the changes to default parameters.

[Martyn Gigg]

Merge remote-tracking branch 'feature/6667_ISISReductionInMantid'

Refs #6667 Conflicts:

Code/Mantid/scripts/CMakeLists.txt Code/Mantid/scripts/Inelastic/DirectEnergyConversion.py

Changeset: c3362412c6241d9929095408b11737fae087cbf0

[Martyn Gigg]

Merge remote-tracking branch 'feature/6667_ISISReductionInMantid'

Refs #6667 Conflicts:

Code/Mantid/scripts/CMakeLists.txt Code/Mantid/scripts/Inelastic/DirectEnergyConversion.py

[Martyn Gigg]

Merge remote-tracking branch 'origin/feature/6667_ISISReductionInMantid'

[Stuart Campbell]

This ticket has been transferred to github issue ​7513