Bug when bands are not in wavelength order

[jgrn307]

I am doing a spectral profile of a USDA NAIP image, which has a weird characteristic in which the bands are not in wavelength order. In microns, they are:

0.65;0.55;0.45;0.85

The issue for this is that the lines connecting the values for each wavelength are following band order, not wavelength order, so they mess up in this case:

Can you fix this by making sure when the x-axis steps are not in order, the line of the plots follows the steps, not the band order?