Tweaks for z>4 (SMF, decrease M/L, allow bluer colors), increase qualilty of UVJ colors and mass limits

This release builds upon the previous version to improve the match of the mock catalog with observations at z>4. In particular, the stellar mass function has now a more reasonable shape (the low mass slope of the SMF was becoming unrealistically steep at z>6), the M/L ratio is manually decreased to reflect the change in SFH, and the UVJ stellar library is extended to cover bluer colors.

Another important improvement concerns the UVJ stellar library. It is now rebuilt by computing the UVJ colors of each observed galaxy from the FAST template itself, rather than using the colors computed by EAZY. There was a slight mismatch between the two, that resulted in a relatively poor sampling of the UVJ space. The filter library now includes rest-frame filters for U and V magnitudes.

The algorithm to estimate the evolving stellar mass limit with redshift (employed when maglim=... is used) is rewritten to take into account the evolution of the typical galaxy color with mass and redshift, instead of assuming the worst case scenario systematically. This will result in a more efficient mass limit such that fewer galaxies will be generated below the requested magnitude cut.

The egg-getsed program can now produce the full SED of the galaxy, rather than just the disk or bulge components separately. Simply avoid specifying the component=... argument to benefit from this behavior. The phy++ library was also fixed to prevent bugs when the number of SEDs stored in the binary file was too large.

Finally, the code is updated to the latest phy++ API.