Fix plot crash on negative errors

[jhaigh0]

Description of work

Fix crash in the case of plotting data with negative errors.

Instead, print a warning and plot without errors.

Fixes #37973

To test:

• Load this data with negative errors
  BCA299_Subt_QLd_Result.zip

• Plot Spectra with errors 6-8

• Should see a warning printed but plot opens normally.

---

Reviewer

Please comment on the points listed below (full description).
Your comments will be used as part of the gatekeeper process, so please comment clearly on what you have checked during your review. If changes are made to the PR during the review process then your final comment will be the most important for gatekeepers. In this comment you should make it clear why any earlier review is still valid, or confirm that all requested changes have been addressed.

Code Review

• Is the code of an acceptable quality?
• Does the code conform to the coding standards?
• Are the unit tests small and test the class in isolation?
• If there is GUI work does it follow the GUI standards?
• If there are changes in the release notes then do they describe the changes appropriately?
• Do the release notes conform to the release notes guide?

Functional Tests

• Do changes function as described? Add comments below that describe the tests performed?
• Do the changes handle unexpected situations, e.g. bad input?
• Has the relevant (user and developer) documentation been added/updated?

Does everything look good? Mark the review as Approve. A member of @mantidproject/gatekeepers will take care of it.

Gatekeeper

If you need to request changes to a PR then please add a comment and set the review status to "Request changes". This will stop the PR from showing up in the list for other gatekeepers.

[Despiix]

Nevermind all good

[Despiix]

Tested according to test instructions and read the code. Works well.

[rbauststfc]

Sorry, a quick question before merging this one in - the linked issue has a comment in the description: "Need a second opinion on this: Do we then attempt to plot the data but taking the absolute positive value of the errors?". This PR has gone with the option to plot the absolute positive value of the errors, and I was just wondering how we came to that final decision?

[jhaigh0]

Sorry, a quick question before merging this one in - the linked issue has a comment in the description: "Need a second opinion on this: Do we then attempt to plot the data but taking the absolute positive value of the errors?". This PR has gone with the option to plot the absolute positive value of the errors, and I was just wondering how we came to that final decision?

I was thinking, a) this is a very rare scenario and b) there is not currently a good mechanism for 'rejecting' a plot; not try catch around the complicated plotting process etc. I thought, since this is a very unlikely problem, the simple option rather than the complex one was better.

[rbauststfc]

Thanks for the explanation @jhaigh0, can definitely see where you're coming from, particularly if it's complicated to reject plotting altogether!

Although this is a rare scenario, what concerns me slightly is if a user could quite easily miss the warning message in the messages pane and then could end up slightly misled about their error values because converting them to positive makes them look like valid errors. I've had a quick chat with Richard to try and get a science perspective on this and his opinion was that if the error is negative then it's garbage and should probably be set to null value (which may be 0 or NaN). We're not sure what happens to NaN values when plotting (perhaps ignored so may look like 0 anyway?), but either way Richard felt that plotting the absolute value of a garbage number was more misleading. What do you think?

[RichardWaiteSTFC]

Further to the above, it is possible that taking the abs of the error gives you the correct result by chance - for example maybe there's a missing abs in the error propagation when the workspace is produced - but in my opinion would this would be 'fixing' the bug in the wrong place and as @rbauststfc says potentially makes it difficult for the user to see something is wrong.

[jhaigh0]

okay understood, I will try to change it to plot without errors + a warning in this case.

[Despiix]

Following the discussion above, I tested based on the coding instructions and it works well. The updated code does not take the the abs of the error anymore.

[rbauststfc]

Thanks @jhaigh0 and @Despiix for taking the time to have a further look at this one. The solution is looking really good to me - I'll get this merged in!