Bug visualizing auto processing COSY

[jobo322]

the source of the problem is the exponential window function activated by default in the apodization panel. It conduces to a NaN in the result of the automatic processing. Is it possible to reproduce this with `cyclosporin/cyclosporin_cosy.zip from bruker-data-test, input filter in FID or preferFID.

it conduce that the slice of contour levels is broken to NaN for the minLevel by default.

we could update the minLevel after each filter modification or set zero if the minLevel got NaN for any reason.

[hamed-musallam]

the source of the problem is the exponential window function activated by default in the apodization panel. It conduces to a NaN in the result of the automatic processing. Is it possible to reproduce this with `cyclosporin/cyclosporin_cosy.zip from bruker-data-test, input filter in FID or preferFID.

Be aware that default1Dapodization is taken from nmr-processing and is currently the same for both 1D and 2D. This setting applies to auto-processing, and if we add a new filter in NMRium, the defaults should be defined according to the experiment

[hamed-musallam]

we could update the minLevel after each filter modification or set zero if the minLevel got NaN for any reason.

I’m not sure about this because updating the contour levels every time we change filters after the FT will override the user’s current settings. If the user has a specific min/max level, this approach would force them to manually reset it each time.

@lpatiny

What do you think should be the behavior in this case?

[targos]

It conduces to a NaN in the result of the automatic processing

This is the bug that should be fixed. We should not workaround NaN. We should avoid it in the first place.

[jobo322]

It conduces to a NaN in the result of the automatic processing

This is the bug that should be fixed. We should not workaround NaN. We should avoid it in the first place.

indeed, but in this case one selection of a wrong filter option conduced to this NaN. it could happening in a another situation, and manually changing the filter option the processing works well, but if the min-max contour level is broken the user would not visualize the result correctly. Can I keep the refactoring, I think it is the most easy and less invasive way to solvent the visualization problem.

[lpatiny]

What is exactly the code that generates a NaN and why ? Try to create a simple test case that generates a NaN at the level of the function that generates this NaN

[jobo322]

it is the first filter apodization passed

after move manually the filter, the structure change

@hamed-musallam could you help me to check where is modified it or where is set the first filter option for apodization

[jobo322]

@hamed-musallam I found the error, it is not in cheminfo/nmrium, sorry. The parameter of apodization from bruker is numerically encoded and 0 is a possible value. It was not expected