1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
|
{ buildPythonPackage
, fetchPypi
, python
, stdenv
, pytest
, glibcLocales
, cython
, dateutil
, scipy
, moto
, numexpr
, pytz
, xlrd
, bottleneck
, sqlalchemy
, lxml
, html5lib
, beautifulsoup4
, hypothesis
, openpyxl
, tables
, xlwt
, runtimeShell
, libcxx ? null
}:
let
inherit (stdenv.lib) optional optionals optionalString;
inherit (stdenv) isDarwin;
in buildPythonPackage rec {
pname = "pandas";
version = "1.0.4";
src = fetchPypi {
inherit pname version;
sha256 = "b35d625282baa7b51e82e52622c300a1ca9f786711b2af7cbe64f1e6831f4126";
};
checkInputs = [ pytest glibcLocales moto hypothesis ];
nativeBuildInputs = [ cython ];
buildInputs = optional isDarwin libcxx;
propagatedBuildInputs = [
dateutil
scipy
numexpr
pytz
xlrd
bottleneck
sqlalchemy
lxml
html5lib
beautifulsoup4
openpyxl
tables
xlwt
];
# For OSX, we need to add a dependency on libcxx, which provides
# `complex.h` and other libraries that pandas depends on to build.
postPatch = optionalString isDarwin ''
cpp_sdk="${libcxx}/include/c++/v1";
echo "Adding $cpp_sdk to the setup.py common_include variable"
substituteInPlace setup.py \
--replace "['pandas/src/klib', 'pandas/src']" \
"['pandas/src/klib', 'pandas/src', '$cpp_sdk']"
'';
setupPyBuildFlags = [
# As suggested by
# https://pandas.pydata.org/pandas-docs/stable/development/contributing.html#creating-a-python-environment
"--parallel=$NIX_BUILD_CORES"
];
disabledTests = stdenv.lib.concatMapStringsSep " and " (s: "not " + s) ([
# since dateutil 0.6.0 the following fails: test_fallback_plural, test_ambiguous_flags, test_ambiguous_compat
# was supposed to be solved by https://github.com/dateutil/dateutil/issues/321, but is not the case
"test_fallback_plural"
"test_ambiguous_flags"
"test_ambiguous_compat"
# Locale-related
"test_names"
"test_dt_accessor_datetime_name_accessors"
"test_datetime_name_accessors"
# Can't import from test folder
"test_oo_optimizable"
# Disable IO related tests because IO data is no longer distributed
"io"
# KeyError Timestamp
"test_to_excel"
# ordering logic has changed
"numpy_ufuncs_other"
"order_without_freq"
# tries to import from pandas.tests post install
"util_in_top_level"
] ++ optionals isDarwin [
"test_locale"
"test_clipboard"
]);
doCheck = !stdenv.isAarch64; # upstream doesn't test this architecture
checkPhase = ''
runHook preCheck
''
# TODO: Get locale and clipboard support working on darwin.
# Until then we disable the tests.
+ optionalString isDarwin ''
# Fake the impure dependencies pbpaste and pbcopy
echo "#!${runtimeShell}" > pbcopy
echo "#!${runtimeShell}" > pbpaste
chmod a+x pbcopy pbpaste
export PATH=$(pwd):$PATH
'' + ''
LC_ALL="en_US.UTF-8" py.test $out/${python.sitePackages}/pandas --skip-slow --skip-network -k "$disabledTests"
runHook postCheck
'';
meta = {
# https://github.com/pandas-dev/pandas/issues/14866
# pandas devs are no longer testing i686 so safer to assume it's broken
broken = stdenv.isi686;
homepage = "https://pandas.pydata.org/";
description = "Python Data Analysis Library";
license = stdenv.lib.licenses.bsd3;
maintainers = with stdenv.lib.maintainers; [ raskin fridh knedlsepp ];
platforms = stdenv.lib.platforms.unix;
};
}
|