[C2E / IDA] - Improve User Input Validation

[Peter Parker]

C2E

Calibration

This has default values read from the Instr Parameter File and are good approximations to start from. However they could need tweaking so we should have checks:

• (PeakMax - Min) > 0.
• BackMax > BackMin - this will also do (BackMax - Min) > 0.
• Perhaps also ((BackMax < PeakMin) OR (BackMin > PeakMax)).

The former is for background on negative side of peak & latter for positive side to ensure background range does not overlap peak.

Resolution

Rebinning has defaults (fixed & those for graphite002) and checks should be like Fury.

• (High-Low) > 0.
• Range multiple of Width.

For background a bit like Calibration:

• End > Start (ensuring (End - Start) > 0).
• For negative side Start > Low.
• For postive side End < High.

Diagnostics

Like Calibration:

• Range1 - (End - Start) > 0.
• If Range2 used - (End-Start) > 0 and ((End2 < Start1) OR (Start2 > End1))

IDA

Elwin

Just like C2E/Diagnostics except no sensible defaults.

CalcCorr

Mainly checks that values are > 0.

• Flat - All Thicknesses > 0.
• Can Angle - cannot be 90 +- 10.
• Cylinder - R2 > R1, R3 > R2 and StepSize < (R2 - R1).
• Sample/Can details - NumberDensity > 0 and (Scattering + Absorption) > 0.
• Sampe & can densities and cross sections cannot be negative.

[Peter Parker]

Refs #5483 - Better validation in Fury.

Accounting for floating point nature of doubles, also Python check.

Changeset: f9834e112bb6a0eda8df612cff2606e4bb04b135

[Peter Parker]

Refs #5483 - Better validation in Fury.

Accounting for floating point nature of doubles, also Python check.

Changeset: f9834e112bb6a0eda8df612cff2606e4bb04b135

[Peter Parker]

Refs #5483 - Changes from Spencer including error checking.

Changeset: dfba8692b30ea7f2adf6c3d57552296476d75008

[Peter Parker]

Refs #5483 - More changes from Spencer re input validation.

Changeset: 5239799c2e104101c6c2d4619b22aa48a7a78190

[Peter Parker]

Refs #5483 - Better validation in Fury.

Accounting for floating point nature of doubles, also Python check.

Changeset: f9834e112bb6a0eda8df612cff2606e4bb04b135

[Peter Parker]

Refs #5483 - Changes from Spencer including error checking.

Changeset: dfba8692b30ea7f2adf6c3d57552296476d75008

[Peter Parker]

Refs #5483 - Add more validation.

• Cover the case when using Plot Raw in Calib/Diag where the user might

accidentally choose a run from a different instrument than the one that is selected in the Energy Transfer tab. This was causing Mantid to crash.

Changeset: a37f2e964ee8d41c8d67df2028ebb4b71dfd2744

[Peter Parker]

Refs #5483 - Add validation to C2E/IDA where missing.

This should cover all the items in the ticket description that haven't already been covered.

Attempted to clean up a lot of the boiler plate code surrounding detection and reporting of invalid inputs.

A couple of the items on CalcCorr may be subject to change.

Changeset: 8b734ddd84a1689e179d419f89fd02e496f5e3a3

[Peter Parker]

Refs #5483 - Delete errant qualification not shown in MSVC.

Changeset: 80c19ec7043cffdb2ecde70d28dae76d676f4b1f

[Peter Parker]

Refs #5483 - Fix another quirk not shown by MSVC.

Changeset: db54d9912f4e50f388f6b58e5e26a63d2d9786fb

[Peter Parker]

Refs #5483 - Final amendments.

Changeset: 4fd4833dcfbeaa821e674f271bbe45955e328b81

[Peter Parker]

To test, simply go to the appropriate interfaces and ensure that the inputs are validated as mentioned in the ticket description.

No need for input files - clicking "Run" on these tabs will display a list of which inputs are not valid.

[Dereck Kachere]

Tests were carried out and passed.

Calibration: all conditions desired were met and no ranges were allowed to overlap. Resolution: an error message came up when the range was not a multiple of the width. Background: all conditions were functional. Diagnostics: similar case; no overlaps were allowed. Clever that user cannot even enter a negative sign for the 'Thickness'

[Peter Parker]

Refs #5483 - Add more validation.

• Cover the case when using Plot Raw in Calib/Diag where the user might

accidentally choose a run from a different instrument than the one that is selected in the Energy Transfer tab. This was causing Mantid to crash.

Changeset: a37f2e964ee8d41c8d67df2028ebb4b71dfd2744

[Peter Parker]

Refs #5483 - Add validation to C2E/IDA where missing.

This should cover all the items in the ticket description that haven't already been covered.

Attempted to clean up a lot of the boiler plate code surrounding detection and reporting of invalid inputs.

A couple of the items on CalcCorr may be subject to change.

Changeset: 8b734ddd84a1689e179d419f89fd02e496f5e3a3

[Peter Parker]

Refs #5483 - Delete errant qualification not shown in MSVC.

Changeset: 80c19ec7043cffdb2ecde70d28dae76d676f4b1f

[Peter Parker]

Refs #5483 - Fix another quirk not shown by MSVC.

Changeset: db54d9912f4e50f388f6b58e5e26a63d2d9786fb

[Peter Parker]

Refs #5483 - Final amendments.

Changeset: 4fd4833dcfbeaa821e674f271bbe45955e328b81

[Stuart Campbell]

This ticket has been transferred to github issue ​6329