Fitting and Graphing Bug

[Owen Arnold]

If you create a fit, and then graph the fit result, then rerun the fit, it deletes the plot. This is easily verified by creating a simple linear background fit.

This has been reported by Max Skoda. What he wants to do is simply see the result of the new fit output.

[Nick Draper]

bulk move to assigned at the into of the triage step

[Owen Arnold]

Users would very much like this fixed for the next release. Max has just been asking me about this again.

[Nick Draper]

verified as occuring on Windows (and propbably everywhere else)

[Nick Draper]

re #8708 remove the explicit delete of workspaces before fitting

Changeset: ae9c22cfc55d5729e9fbaace046eff6c894c4924

[Nick Draper]

Must be tested by Anders

This ticket reverses a change that was part of a large number of changes made in ticket #3790 - Fitted workspaces within the muon interface are grouped.

Done by Robert Whitely in 2011. The whole approach to grouping has been refactored since and it is likely that this code is not needed any more.

However it needs carefull testing specifically using the muon interface, but I will also give a test case here for the general case.

General Case:

1. load any workspace, create a plot "plot a" and fit the data (use a linear background or something)
2. Create a plot "plot b" from the plot output, the inputdata_workspace workspace.
3. Run the fit again after moving the fitting boundary.
4. The graph should "plot b" should update, rather than disappaearing.

Muon case:

1. You know this better than I, but you need to test out the moun fitting in particular.

[Anders Markvardsen]

Merge remote-tracking branch 'origin/feature/8708_graph_windows_vanish_on_refitting'

Full changeset: ​880bf72b54f444e50364bd88f10351df7a11e607

[Anders Markvardsen]

Have tested muon interface, and the general test. Good news it that I did not found any changes where it did not work.

[Stuart Campbell]

This ticket has been transferred to github issue ​9552