paris-saclay-cds / ramp-workflow

Toolkit for building predictive workflows on top of pydata (pandas, scikit-learn, pytorch, keras, etc.).
https://paris-saclay-cds.github.io/ramp-docs/
BSD 3-Clause "New" or "Revised" License
68 stars 42 forks source link

Bad version #191

Closed GuillaumeDesforges closed 4 years ago

GuillaumeDesforges commented 4 years ago

Hello,

Bug report

I installed the package from GitHub using Nix. I get the following:

% python                                                                                                          ~
Python 3.7.5 (default, Oct 14 2019, 23:08:55) 
[GCC 8.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import rampwf
>>> rampwf.__version__
'0+unknown'

More information

Nix overlay

let
  customPythonPackages = self: super: rec {
    python3 = super.python3.override {
      packageOverrides = selfpy: superpy: {
        ramp-workflow = superpy.buildPythonPackage rec {
          pname = "ramp-workflow";
          version = "v0.2.0+53.gf48b024";

          doCheck = false;

          propagatedBuildInputs = with superpy; [
            numpy scipy pandas scikitlearn cloudpickle click
          ];

          src = super.fetchFromGitHub {
            owner = "paris-saclay-cds";
            repo = "ramp-workflow";
            rev = "f48b024642c0756d9e12717067cbca224e34b9ce";
            sha256 = "1xlw3jyfn3nkfbryj8dcm04h5l82w2bhds8cynik7v6mds4jxgll";
          };

          meta = with superpy.stdenv.lib; {
            homepage = "https://github.com/paris-saclay-cds/ramp-workflow";
            description = "Define, build, manage, and optimize data analytics workflows.";
          };
        };
};
    };

    pythonPackages = super.recurseIntoAttrs (python3.pkgs);
  };
in
  [
    customPythonPackages
  ]
agramfort commented 4 years ago

@glemaitre @lucyleeow @kegl we have a running version mechanism now based on git hash. It creates the issue. We should have more common version numbers and rampwf should have a fixed version on pypi ideally