Add algorithm to truncate POLDI spectra according to chopper speed

[Michael Wedel]

The number of time bins in POLDI data is connected to the chopper rotation speed and the bin width, which usually results in 500 bins. The actual raw data contains more than those 500 bins as a check if everything is okay with the measurement. The 501. bin should not contain more than a few counts and the 502. bin should not contain anything.

The algorithm I want to add will check these bins in the raw data after loading and sum the counts over each spectrum for each of the additional bins and output this in a table workspace (optional). Then, it will truncate the data to the appropriate number. This is important for later data analysis, especially calculation/fitting of 2D-spectra.

[Michael Wedel]

Refs #9630. Created PoldiTruncateData algorithm

Implemented the basic algorithm with error handling for workspaces with the wrong size. Unit tests are available.

Changeset: 1f7e3be796fd3735f21e291e4e1a5eb058e9375c

[Michael Wedel]

Refs #9630. Added documentation to cpp file.

Changeset: cb9b6463762f7c6b5b3755b3f4515bd3f6ccf644

[Michael Wedel]

Refs #9630. Added more unit tests.

Changeset: 98eca02af42a81eef353c0a6dad4217a35abcb2f

[Michael Wedel]

Refs #9630. Added documentation and usage examples.

The usage examples will fail until master has been merged into this branch as well, after #9589 has been closed.

Changeset: 253b484f039fd7e1537e72b13bf01093e64eeeba

[Michael Wedel]

Refs #9630. Added system test for PoldiTruncateData

The test goes through three possible use cases: Too many time bins (the usual raw POLDI data), exactly correct number of bins and too few time bins.

Changeset: fa16c9f2ec7507e32d7d529680baf97724a246e8

[Michael Wedel]

Testing information:

There is a new Algorithm named PoldiTruncateData. The usage example scripts in the documentation show how it should work in the "happy case". If you like, you can create a color fill plot of the raw and the processed data - in the raw data you should see a stripe on the right hand side of the spectrum, which does not contain any counts. This is stripe is gone in the processed data.

Truncating the workspace manually with CropWorkspace before running PoldiTruncateData will lead to different results, depending on how much you cut off. This is covered by the system tests. If the workspace is too small, it will not create an output workspace, but generate error messages instead.

[Michael Wedel]

Refs #9630. Some minor clean-up.

Removed using namespace directive from header file, removing output property on timing problems, added initialization unit test.

Changeset: be6a5fb0301a10a2f536bc070717cf663ab0211f

[Michael Wedel]

Since #9589 is closed now, the required data files for usage tests are in master now, so I am marking this ticket as fixed as well.

[Nick Draper]

Merge remote-tracking branch 'origin/feature/9630_truncate_poldi_data'

Full changeset: ​0749f65ce926e3ab212f87185bbb9a466d44f147

[Michael Wedel]

Merge branch 'feature/9630_poldi_truncate_data_test' into feature/10074_merge_poldi_system_tests

Full changeset: ​45cdc5b1ed15c441bd1272da76b5c152088667f4

[Martyn Gigg]

Merge branch 'feature/9630_poldi_truncate_data_test'

Full changeset: ​59c2f345040c5c1f27c8d977a50d713187dd3568

[Stuart Campbell]

This ticket has been transferred to github issue ​10473