Michael Walther

merged master back into here to enable the changes from v9, please double check... removed the "unblind" option. Not sure if that was from the starting point of this PR...

sounds like a plan to me

Limits for coveralls are set to fail when coverage decreases by > 1%. So this is the expected behaviour given a 1.2% change in coverage. [The coveralls output](https://coveralls.io/builds/66830898) shows that...

in case the current analysis routines are expected to already be able to do things properly (sry, didn't check), it might be worth changing the order (also so that the...

One question regarding the coadding: shouldn't one do co-adding for continuum fitting always? Or at the very least make sure that the linear quasar continuum rescaling fit to each object...

The issue is that it might take a lot of changes in how I/O or cont fitting is done. Maybe it would be useful to have a look at the...

Ok, I think what needs to be done to make the coverage go up is having tests analogous to add tests like [the tile tests in data_tests](https://github.com/igmhub/picca/blob/eecadaa92d1acdf8a78f55814d74d5d41d1f351f/py/picca/tests/delta_extraction/data_tests.py#L1163) with config settings...

I added a skeleton for the data readin from tiles. Lets see how the coverage reacts

Ok, testing multiple ways of the tile based non-coadded analysis (i.e. with/without keeping individual spectra and one vs multiple CPUs) actually makes the coverage test pass, nevertheless would be good...

I think given the biases the "best exp" case might not be a good one to keep and having the 2 options of either using individual exposures for cont fitting...