956 data in cut plot distorted after ticking and unticking log scale in x axis

[GuiMacielPereira]

Description of work:

• Fixed bug where data gets distorted from setting log scale to x axis and then setting it to linear again
• Restructured section of the code dealing with updating the log scales on the axes.

To test:

• Open cut plot with some example data
• Open plot options > tick Logarithm for x axis > 'OK'
• Check log scale is applied to x axis
• Open plot options again > untick Logarithm for x axis > 'OK'
• Check plot is back to original format
• Repeat the same steps for y axis
• Now put both axis on a linear scale
• Go to Information > Bragg Peaks > check Aluminium
• Change y axis to log scale
• The lines for the Aluminium bragg peaks should stay symmetrical around zero
• Change y axis to linear
• The lines for the Aluminium bragg peaks should stay symmetrical around zero

Fixes #956

[robertapplin]

Thanks for the changes. Just did some functional testing and I get an error specifically for this part of the instructions:

Go to Information > Bragg Peaks > check Aluminium Change y axis to log scale

Traceback (most recent call last):
  File "C:\Users\mlc47243\Documents\mantid-dev\mslice\src\mslice\presenters\plot_options_presenter.py", line 126, in get_new_config
    setattr(self._model, key, value)
  File "C:\Users\mlc47243\Documents\mantid-dev\mslice\src\mslice\plotting\plot_window\cut_plot.py", line 675, in y_log
    linthresh_val = _gca_sym_log_linear_threshold(axis_data)
  File "C:\Users\mlc47243\Documents\mantid-dev\mslice\src\mslice\plotting\plot_window\cut_plot.py", line 36, in _gca_sym_log_linear_threshold
    axis_data = np.array(axis_data)
ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (2,) + inhomogeneous part.

[GuiMacielPereira]

My goodness, well spotted. I'll fix that.

[GuiMacielPereira]

• Fixed it. This happened because I didn't consider the case where converting a list of arrays with different dimensions into a single array fails when using np.array(). Replaced it by np.concatenate(), as it's not important to preserve the dimensions.