Enhancements to rocking curve algorithm

[Russell Taylor]

Following on from #6150, which provided a quick initial implementation, further development of the SumEventsByLogValue algorithm is required. This will be quite a change in that the output will become a table workspace containing the total counts as a function of log value, but also containing the average value of other time-series logs as a function of the requested log value - either for any input log or specifically for the (monotonically increasing) scan_index log.

Another requirement is the ability to carry out some kind of normalisation (e.g. by monitor, time or proton charge) or the retention of the information required so that one can be done later.

[Russell Taylor]

Re #6314. Add protection against an empty log being used as input.

Done via the introduction of a validateInputs method and tests.

Changeset: c402c477b5d0b45d5f04d919635cbe1be122e974

[Russell Taylor]

Re #6314. Change integer logs to output results in a TableWorkspace.

Can alternatively have them in a (histogram) Workspace2D if binning parameters are given (i.e. treating them the same as double TSPs).

Changeset: a5ec419aa738f5ec67a5718af98fe9c236b76d9b

[Russell Taylor]

Re #6314. Add a method to calculate the time-averaged value of a log

...in a given range of time(s). This is needed for the SumEventsByLogValue algorithm.

Changeset: a3c1b823bf71ea160b570ad7f7b8e01fb5f3fde0

[Russell Taylor]

Re #6314. Add columns for normalisation & other logs.

Columns for nomalisation are: monitor counts, proton charge & time. Columns will also be added to give the average value of other time-series logs for each step.

Changeset: 234552b1cbf5a362dc037967166dcade323b4d93

[Russell Taylor]

Re #6314. Fix Windows builds.

Changeset: d7325746d60f4e321c93fea6611da21b7bc3cb2f

[Russell Taylor]

Re #6314. Add protection against an empty log being used as input.

Done via the introduction of a validateInputs method and tests.

Changeset: cb5b85e2b9f607021be7dbb651b50b27aa1b3034

[Russell Taylor]

Re #6314. Change integer logs to output results in a TableWorkspace.

Can alternatively have them in a (histogram) Workspace2D if binning parameters are given (i.e. treating them the same as double TSPs).

Changeset: 7b646e2419d7d9b21b13c737653d2c9b7419761e

[Russell Taylor]

Re #6314. Add a method to calculate the time-averaged value of a log

...in a given range of time(s). This is needed for the SumEventsByLogValue algorithm.

Changeset: fff6fc0f1970e72041434d054c80706ce35bf676

[Russell Taylor]

Re #6314. Add columns for normalisation & other logs.

Columns for nomalisation are: monitor counts, proton charge & time. Columns will also be added to give the average value of other time-series logs for each step.

Changeset: b4a15de6148f6fb0e15e7ec5392364cbdbc16294

[Russell Taylor]

Re #6314. Fix Windows builds.

Changeset: 58e14f1ef7a96bb28766a17ddebb9f3c1f69de60

[Russell Taylor]

Re #6314. Add a method to calculate the time-averaged value of a log

...in a given range of time(s). This is needed for the SumEventsByLogValue algorithm.

Changeset: a3c1b823bf71ea160b570ad7f7b8e01fb5f3fde0

[Russell Taylor]

Re #6314. Add columns for normalisation & other logs.

Columns for nomalisation are: monitor counts, proton charge & time. Columns will also be added to give the average value of other time-series logs for each step.

Changeset: 234552b1cbf5a362dc037967166dcade323b4d93

[Russell Taylor]

Re #6314. Fix Windows builds.

Changeset: d7325746d60f4e321c93fea6611da21b7bc3cb2f

[Russell Taylor]

Re #6314. Make performance test take longer.

Changeset: e5e3fbced45a401e5864e2f6936c15fb651482ca

[Russell Taylor]

Re #6314. Add a performance test for CropWorkspace (for events).

Changeset: f71ed34162d559124ac8946e0d7ead8403ef1f16

[Russell Taylor]

Re #6314. Documentation added.

Changeset: b8c338cb9a7dbc16631b8d147e098fc725388de3

[Russell Taylor]

Re #6314. Add a workflow algorithm to generate a rocking curve.

Works off of the scan_index log variable that is present in the ADARA stream. Uses SumEventsByLogValue underneath and also adds the possibility to mask pixels and narrow the X range.

Changeset: e9ada7264b750e2fc9ed76196b955e286e9d792b

[Russell Taylor]

Re #6314. Try to fix test on Windows & Ubuntu.

Changeset: 185ac2e423f159a363f6427e29a80ec44065b7e4

[Russell Taylor]

Re #6314. Fix unit test on Ubuntu.

Because nothing from Algorithms was explicitly used in any test (only via the child algorithm mechanism), the library wasn't being linked to which led to complaints that the FilterByXValue child algorithm wasn't registered. I added some dummy code to force the linking.

Changeset: d2f68fb3d815a0c91bd1d17839588285639ec842

[Russell Taylor]

Re #6314. Use binary, not linear, search in EventList::maskTofHelper.

Changeset: d8c39456744c64cb7a3108cef5ebb30388628e5c

[Russell Taylor]

Re #6314. Make performance test take longer.

Changeset: c23d38ad5831477035b2490299d3a747f15db3ba

[Russell Taylor]

Re #6314. Add a performance test for CropWorkspace (for events).

Changeset: 14c3be93440c0ee45cbbf0ac51213ae3d920440e

[Russell Taylor]

Re #6314. Documentation added.

Changeset: 527891b7e4d97b1348fa294eff4041d4301bcde0

[Russell Taylor]

Re #6314. Add a workflow algorithm to generate a rocking curve.

Works off of the scan_index log variable that is present in the ADARA stream. Uses SumEventsByLogValue underneath and also adds the possibility to mask pixels and narrow the X range.

Changeset: 69b19a5d0a1ad2896d3950f282209dfe4e3dc8fc

[Russell Taylor]

Re #6314. Try to fix test on Windows & Ubuntu.

Changeset: b9e4b421fafc02e82c1c3202f3ddcb7ed034a4ed

[Russell Taylor]

Re #6314. Fix unit test on Ubuntu.

Because nothing from Algorithms was explicitly used in any test (only via the child algorithm mechanism), the library wasn't being linked to which led to complaints that the FilterByXValue child algorithm wasn't registered. I added some dummy code to force the linking.

Changeset: caea696d89bd3b95108befda2974bfdafdaef518

[Russell Taylor]

Re #6314. Use binary, not linear, search in EventList::maskTofHelper.

Changeset: e30e6fe3f7c25a97dfbb7c2ece5e27a266b37a29

[Russell Taylor]

Re #6314. Parallelize & add progress reporting.

Changeset: 9bda6e0515e21dd3859cd2efea4dd6260d526185

[Russell Taylor]

Re #6314. Restrict to workspaces holding unweighted events.

Changeset: 7cfcf296537dbbecaa44b3f83150128694f5425e

[Russell Taylor]

Re #6314. Documentation.

Changeset: 6950ced3545459172a9943fb5330a5538e0995f3

[Russell Taylor]

Re #6314. Doxygen documentation.

Changeset: f7ab74b1f383c79c2facee0df758ab14cd2f8968

[Russell Taylor]

Tester: The algorithms to look at are SumEventsByLogValue and RockingCurve and they're relatively simple to use so the wiki documentation should suffice for directions. The former can be used with many data files, but for the latter you'll probably need to use HYSA_2934 from the system tests repo. I'm going to add a system test for it, but need to fix CheckWorkspacesMatch to work with table workspaces first.

The bit about normalisation in the description has been fulfilled by including columns in the table. These will then be manipulated by the Rocking Curve custom interface (which isn't included in the menus in this release).

There are a few TODOs in the source code - these will be addressed in the next iteration.

[Alex Buts]

The SumEventsBylogValue work though not as I expected it to work. This is probably what it should do but I have expressed my expectation in low priority ticket #6494 (In case somebody else have similar expectations)

Rocking curve is much work in progress but what I was able to test works fine too. Of course, more tests for this algorithm are needed and planned.

[Stuart Campbell]

This ticket has been transferred to github issue ​7160