GitLab

Updating the parametrizations to be more complete.
Changing the data object to handle cosine correction.

Fitting of diffuse with cosine weighting

 * COS_TLV: the retro cosine
 * COS_TLR: the classic reflected and light cosine

The parametrization is still an issu, it should be converted to an object where
the dimension is included so that the convertion can be done.

Cleaning the parsing code for Beckmann and retro Beckmann lobes.

They should not call the get_name function

Correction of the L2 & Linf distance with respect to the objects params
Updating the example in python

The Beckmann distribution wasn't working due to the lack of correct Jacobian for the Geometric term. I commented the Geometric term from the BRDF to get correct fit. Will need to work on that.

Allowing to extract slices of data from the data2gnuplot interface

to the loading. add --data-correct-cosine in the command line

I need to debug the back projection to see if it works

Changing the bootstrap of fresnel to be in the middle of the domain.

Compound with product bootstraping. It does not solve the issue when the bootstrap is done with more depth to the compound tree.

The parallel algorithm runs, but I need to check if it provides the correct answer