better resiliance of bift on bad data

[kif]

tested on unsubtratedd SANS data

[maaeli]

Would it be possible to add the SANS data to the test data? Might be worth making a more permanent test out of it...

[kif]

I am in contact with Anne to get access to those data. Apparently, the test dataset was not background corrected thus the crash.
Overall, it is better to make it resilient to those crashes.

[kif]

This turns out to be not that simple ...

• either one rebins SANS data to regular and fine qbins and fall back on the SAXS analysis
• or one has to restart with I(q) = sum P(r)*sinc(qr) dr and re-implement them BIFT ...