depot/third_party/nixpkgs/pkgs/development/python-modules/datashader/default.nix
Default email fcc03dbfc5 Project import generated by Copybara.
GitOrigin-RevId: ff96a0fa5635770390b184ae74debea75c3fd534
2021-02-13 15:23:35 +01:00

90 lines
1.6 KiB
Nix

{ lib
, buildPythonPackage
, fetchPypi
, fetchpatch
, dask
, distributed
, bokeh
, toolz
, datashape
, numba
, numpy
, pandas
, pillow
, xarray
, colorcet
, param
, pyct
, pyyaml
, requests
, scikitimage
, scipy
, pytest
, pytest-benchmark
, flake8
, nbsmoke
, fastparquet
, testpath
, nbconvert
, pytest_xdist
}:
buildPythonPackage rec {
pname = "datashader";
version = "0.11.1";
src = fetchPypi {
inherit pname version;
sha256 = "b1f80415f72f92ccb660aaea7b2881ddd35d07254f7c44101709d42e819d6be6";
};
patches = [ (fetchpatch {
# Unpins pyct==0.46 (Sep. 11, 2020).
# Will be incorporated into the next datashader release after 0.11.1
url = "https://github.com/holoviz/datashader/pull/960/commits/d7a462fa399106c34fd0d44505a8a73789dbf874.patch";
sha256 = "1wqsk9dpxnkxr49fa7y5q6ahin80cvys05lnirs2w2p1dja35y4x";
})];
propagatedBuildInputs = [
dask
distributed
bokeh
toolz
datashape
numba
numpy
pandas
pillow
xarray
colorcet
param
pyct
pyyaml
requests
scikitimage
scipy
testpath
];
checkInputs = [
pytest
pytest-benchmark
pytest_xdist # not needed
flake8
nbsmoke
fastparquet
pandas
nbconvert
];
# dask doesn't do well with large core counts
checkPhase = ''
pytest -n $NIX_BUILD_CORES datashader -k 'not dask.array'
'';
meta = with lib; {
description = "Data visualization toolchain based on aggregating into a grid";
homepage = "https://datashader.org";
license = licenses.bsd3;
maintainers = [ maintainers.costrouc ];
};
}