math/py-numdifftools: New port: Solver of automatic numerical differentiation problems
This commit is contained in:
@@ -1068,6 +1068,7 @@
|
||||
SUBDIR += py-nevergrad
|
||||
SUBDIR += py-nlopt
|
||||
SUBDIR += py-numba-stats
|
||||
SUBDIR += py-numdifftools
|
||||
SUBDIR += py-numexpr
|
||||
SUBDIR += py-numpoly
|
||||
SUBDIR += py-numpy
|
||||
|
||||
@@ -0,0 +1,30 @@
|
||||
PORTNAME= numdifftools
|
||||
DISTVERSION= 0.9.42
|
||||
CATEGORIES= math python
|
||||
MASTER_SITES= PYPI
|
||||
PKGNAMEPREFIX= ${PYTHON_PKGNAMEPREFIX}
|
||||
|
||||
MAINTAINER= yuri@FreeBSD.org
|
||||
COMMENT= Solver for automatic numerical differentiation problems
|
||||
WWW= https://github.com/pbrod/numdifftools/
|
||||
|
||||
LICENSE= BSD3CLAUSE
|
||||
LICENSE_FILE= ${WRKSRC}/LICENSE.txt
|
||||
|
||||
BUILD_DEPENDS= ${PYTHON_PKGNAMEPREFIX}pdm-backend>0:devel/py-pdm-backend@${PY_FLAVOR}
|
||||
RUN_DEPENDS= ${PYTHON_PKGNAMEPREFIX}numpy>=1.21.2:math/py-numpy@${PY_FLAVOR} \
|
||||
${PYTHON_PKGNAMEPREFIX}scipy>=1.7.3:science/py-scipy@${PY_FLAVOR}
|
||||
TEST_DEPENDS= ${PYTHON_PKGNAMEPREFIX}hypothesis>=3.6:devel/py-hypothesis@${PY_FLAVOR} \
|
||||
${PYTHON_PKGNAMEPREFIX}statsmodels>=0.6:math/py-statsmodels@${PY_FLAVOR}
|
||||
|
||||
USES= python
|
||||
USE_PYTHON= autoplist pep517 pytest
|
||||
|
||||
NO_ARCH= yes
|
||||
|
||||
TEST_ENV= ${MAKE_ENV} MPLBACKEND=Agg PYTHONPATH=${STAGEDIR}${PYTHONPREFIX_SITELIBDIR}
|
||||
PYTEST_ARGS= --override-ini=addopts= src/numdifftools/tests/
|
||||
|
||||
# tests as of 0.9.42: 116 passed, 28 skipped, 242 warnings in 12.80s
|
||||
|
||||
.include <bsd.port.mk>
|
||||
@@ -0,0 +1,3 @@
|
||||
TIMESTAMP = 1780681268
|
||||
SHA256 (numdifftools-0.9.42.tar.gz) = 866675171f293c4bf2f1e1c5bf9b88a07d5396903e3b3e7fcc3879e2a01cfbc1
|
||||
SIZE (numdifftools-0.9.42.tar.gz) = 79612
|
||||
@@ -0,0 +1,5 @@
|
||||
Numdifftools is a suite of tools written in Python to solve automatic
|
||||
numerical differentiation problems in one or more variables. It
|
||||
provides functions for calculating derivatives, Jacobians, Hessians,
|
||||
and gradients, all with error estimates. The methods are based on
|
||||
the complex-step derivative approximation and extrapolation algorithms.
|
||||
Reference in New Issue
Block a user