Separate Property manager from Direct inelastic reduction script

[Alex Buts]

This is the part of agreed work for unifying ISIS direct reduction scripts.

The class responsible for ISIS reduction should be separated into property managed and executive part.

This will also make it much close to SNS reduction bearing in mind possible merging them in a future.

[Alex Buts]

Re #10532 Initial separation

Changeset: 072b1f1be1660ffd62b8113f1ce88960c4dc1746

[Alex Buts]

Re #10532 Further small steps

Changeset: 7fce5d29b646f67dce972fc89f2f32cc082a14e5

[Alex Buts]

refs #10532 Defined generic operations over reduction parameters.

Changeset: 7782b6888550995b456220d9bb83f35ab7d8d81d

[Alex Buts]

refs #10532 More steps in defining property manager

Changeset: 789b2ddb0faa0c35cba452d0698b9221d1066e33

[Alex Buts]

refs #10532 Initial property manager code

Changeset: f001162c880ac336378e5f938e1c8d5cd8858329

[Alex Buts]

refs #10532 reimplemented parts of the property manager

using descriptors

Changeset: a8663ac4f015b06014be10aa14cdfa8a41dd7e54

[Alex Buts]

refs #10532 New class ComplexProperty and started binging

simple and complex reduction properties.

Changeset: aacd44fc6386ccbac834192760a76417d715dfd6

[Alex Buts]

refs #10532 Main functionality moved to separate property manager

Changeset: 77f4ab6d1c3c438ca521b60ecb254117ae1074de

[Alex Buts]

refs #10532 Old interface added for verification

necessary for transition period

Changeset: 13f365043eead1e56512b6e6e25cc4674b5c1b75

[Alex Buts]

refs #10532 minor comments

Changeset: fac2fe30167cb5fba0a618aa074b48e07b3be136

[Alex Buts]

Re #10532 DirectEnergyConversion modified to use DirectPropertyManager

Changeset: 6b99c42a0f34f3ae0c9edf49e34fa36efcdf1f72

[Alex Buts]

refs #10532 further separation in DirectEnergyConversion.

Changeset: 4ee842466af2c23b03057ad4b67cec9296db5ada

[Alex Buts]

Re #10532 minor changes related to changes in diagnose interface

Changeset: f645a8beb50d6808449e1749b12f5c550eaf74f8

[Alex Buts]

refs #10532 Intermediate checkout

walking through DirectEnergyConversion and using PropertyManager everywhere

Replaced all in diagnostics but need to deal with wb integrals next.

Changeset: 9725d89e2d129f4f48e0dea8da8b2ef2699dcafa

[Alex Buts]

Re #10532 Modified and briefly tested do_white and diagnose methods

in DirectEnergyConversion

Changeset: d2f44cd8dcd1bded13efc8a746d00ea2d60f36d7

[Alex Buts]

Re #10532 Fixed diagnose system test

Changeset: 4d4352c39e0ec1c54a249264b213c2838ca53aec

[Alex Buts]

Re #10532 Formally started to run.

system tests are still failing

Changeset: b7e120f1d5a97ac71934960d1ccdad0169e60a46

[Alex Buts]

Re #10532 Minor changes to run separated reduction version.

Changeset: b86b9552493602351370921efdcd1160254cfba3

[Alex Buts]

Re #10532 separated DirectReductionProperties into two

as it become too big to be convenient.

Changeset: 64fb148da3a5bc49b68c18aeed5b2ab220530801

[Alex Buts]

Re #10532 written and tested option to write changed properties

to HDD and load them back.

Changeset: 66e5eedf6756b2cf3714b791e13ec93a2da46a17

[Alex Buts]

Re #10532 Test version of ReductionWrapper (will be abstract and mod

later)

Changeset: 3499986aae721a4f9dc164c889dad61eb32267e9

[Alex Buts]

Re #10532 minor bug fixes.

Changeset: 8aa5c3444ce1a6caeb8ccb37512b88df75ab7bfc

[Alex Buts]

Re #10532 Changing in default log levels and better logging

Changeset: 1ee8235ebe22f619ee3e0fa1061d28e096903e42

[Alex Buts]

Re #10532 fixed simple tests for MARI

and modified ReductionWrapper according to Marcus suggestions (minor)

Changeset: 0b13082d61f22aff7c9dc3834f51b1b3fc5cdf16

[Alex Buts]

Re #10532 Fixing system test for summation

and fixing long standing error in dgreduce (thanks good nobody used internal reduction procedure)

Changeset: cea7d0fed02e813df070d79fe2677debf12190c8

[Alex Buts]

Re #10532 Fixing system tests for summation

and amendments to other tests for Mari appearing due to WB integrals done differently.

Changeset: 71ac4c6ba347fe94cab522d7e32aa4a17f5dc683

[Alex Buts]

Re #10532 fixed system tests for MARI

Changeset: 28985db02c0843e4aa3b2c363dd1dc0813fb8bdb

[Alex Buts]

Re #10532 Fixed Merlin and Maps system tests

Modified property managed to allow it MonovanIntegrationRange accepting absolute energy range (used by MAPS system test)

Changeset: c573b4853895f51b657050668e9ad984839387fe

[Alex Buts]

Re #10532 Minor modifications to fix system tests

Changeset: 98ed67351985876f50cafeab27b774a533a63252

[Alex Buts]

Re #10532 all working, all fixed.

Changeset: bcc791341cbf8ec1e2cd639d4c582033183eed58

[Alex Buts]

Re #10532 changes to fix ISIS system tests

Changeset: e05b8aa04e7937ecbb01f30d5fb6dca4e4122d38

[Alex Buts]

Re #10532 updated SNS code (not if anybody uses it anyway)

Changeset: becaea1a07d4e64a112ab037771def0b1be22680

[Alex Buts]

The changes are both in Mantid and System test branch and are python only.

Difficult ticket to test in details if you are not familiar with ISIS direct inelastic and nobody probably are well enough and the ticket deals with the re-factoring of the code not introducing any new functionality.

But not much testing is needed as doing it I have intentionally not introduced any new functionality (except fixing couple of bugs, which nobody used anyway) but just separated three years of subsequent maintenance changes to interface of ISIS direct inelastic (which caused initial file to grow from ~1K rows to ~3K) into separate classes.

The interface classes are written following the description, outlined in Lutz; Learning Python 2011, mainly chapters 37-38.

The changes are intended to made interface dependent on Instrument_Parameters.xml file only and prohibit access and creation of any properties, not described there.

Doing that I have also tried to maintain and support all functionality, introduced during number of years of maintaining these scripts -- just separated the interface from the code responsible for the reduction itself.

This is mainly successful, though number of small ticket is still necessary to polish the changes.

The proof that the changes work are system tests, verifying the reduction validity.

Tester welcome to look through PropertyManager which is mainly complete but should not bother with DirectEnergyConversion, as this one has been intentionally changed as small as possible to provide all previous functionality.

[Alex Buts]

Re #10532 Minor comments

Changeset: 7c0908a5f8367b490821ca365c9b87b36ea4714c

[Martyn Gigg]

Fix LoadInstrument usage test

The parameter count has changed. Refs #10532

Changeset: 86331033884bb7b6d688df1db2b1c49b32edbc6f

[Martyn Gigg]

After the mass build breakages I'm trying to get them green.

I've fixed the LoadInstrument usage test, so please update your local copy before merging to master.

[Martyn Gigg]

Please don't merge this yet.

There are two usage test failures:

• GetEi
• DgsReduction

See ​http://builds.mantidproject.org/view/Wall%20display/job/develop_doctest/lastCompletedBuild/testReport/

It looks to me that the parameter names for the Ei monitor spectra have been changed but GetEi2 is still looking for the old ones. It also looks like the changes to the parameter names have NOT been applied to SNS instruments so that GetEi fails there too.

Is there any need to change the parameter names?

[Alex Buts]

This ticket does not touch anything in C++ and I do not see why it should.

Does it picks up monitor's parameters names from IDF? A unit test in C++ should verify this feature in such case.

Parameter names have changes to allow assignment from Python by the same names and python does not support minus in properties. There is way to substitute names in Python, so I am reverting change to monitor names, but separate changes are necessary to make this feature more obvious.

[Alex Buts]

Re #10532 Renamed ei monitors names to names with dashes.

Changeset: 7228c3b7000bfa83ac207fec347cfbfa8b552c81

[Alex Buts]

Re #10532 Change to unit tests caused by changes in properties

Changeset: 4d20e6a671b31b75ddd27fe7d6579df7a911069b

[Alex Buts]

Re #10532 Fixed conflicts with develop

Merge branch 'feature/10532_separatePropertyManager' into develop

Conflicts:

Code/Mantid/scripts/Inelastic/DirectReductionProperties.py

Changeset: ea828d464efbb1c1c91a0cefe6db8dae4736bf26

[Alex Buts]

Re #10532 modified deltaE-mode to have dash rather then _

Changeset: 6672970d576d84f88ac48ae952a48575e0f80c77

[Ian Bush]

Merge remote-tracking branch 'origin/feature/10532_separatePropertyManager'

Full changeset: ​c4ada99cdce54ceba0e571bd37d14a6b48e88519

[Ian Bush]

Merge remote-tracking branch 'origin/feature/10532_separatePropertyManager'

Full changeset: ​e7e67016f963645a0620d0545e19b33f82f03f9b

[Stuart Campbell]

This ticket has been transferred to github issue ​11374