search

LupoA / lsdensities

Smeared spectral densities from lattice correlators
GNU General Public License v3.0
3 stars 0 forks source link

Develop #52

Closed LupoA closed 3 months ago

LupoA commented 4 months ago
  1. Address #51
  2. Address #50
  3. Address #18 by adding new algorithmical parameter "plateau_id". It is defaulted to the value that was previously hardcoded.
  4. Removed a bunch of classes that are no longer needed (HLT class and GPerrorHLT class, their job is done by InverseProblemWrapper). Related examples are also removed.
LupoA commented 4 months ago

Also going for #28

LupoA commented 4 months ago

Also #32

the parser now accepts a loglevel option. By default this is set to "warning", so that InverseProblemWrapper doesn't vomit thousands of lines unless request by setting loglevel to "info".

LupoA commented 3 months ago

Does it work for you to have this merged or would it interfere with the release? @nickforce989 @edbennett

nickforce989 commented 3 months ago

@LupoA , at a first glance: It doesn't look like so. Moreover my data release is based on the version that I have uploaded on 'pip', which is the one that is currently on main (before your PR). Unless you don't plan to push it to pip surely it's fine. Otherwise, if you want to push the updated version to pip, I might want to do all the checks to make sure.

edbennett commented 3 months ago

Since Niccolò is pinning against a version from PyPI, then even if there are breaking changes, provided you also update the version number, then it will not cause issues. (Definitely don't push a new version to PyPI without updating the version number, as otherwise it gets very confusing when two different installations of the "same version" have different functionality.)

nickforce989 commented 3 months ago

Ed is right, I was about amending my message saying the same thing.