MAINT: Modify setup requirements#7993
Conversation
a14daf2 to
c32f8f3
Compare
|
This pull request introduces 1 alert when merging c32f8f3 into 19db796 - view on LGTM.com new alerts:
|
fed9d39 to
b65150c
Compare
Remove secondary check which is incompatible with using pip and oldest-numpy-version closes statsmodels#7992
Follow NEP and drop 3.7
b65150c to
d83c175
Compare
| # These are strictly installation requirements. Builds requirements are | ||
| # managed in pyproject.toml | ||
| INSTALL_REQUIREMENTS = { | ||
| "numpy": "1.18", # released December 2019 |
There was a problem hiding this comment.
@bashtage We were having issues using statsmodels with Python 3.9 on a Red Hat cluster due to the install requirement on numpy 1.17 (see numpy/numpy#14147). Glad to see this has been upgraded to numpy 1.18. Any plans for a pypi release in the near future that would include this change?
There was a problem hiding this comment.
Can you not set the c99 flag and build with that in the environment? I'll see if I can cut a 0.13.2 this week. It is overdue a bit.
There was a problem hiding this comment.
Normally we would not bump install req for a released version. These changes are for the 0.14 release.
There was a problem hiding this comment.
Thanks for the quick response! We can use the c99 flag, but our environment makes it difficult to do so cleanly. If this change is not coming until 0.14, we may delay using statsmodels in combination with Python 3.9 until then.
There was a problem hiding this comment.
I suspect that it won't be out until 2014, although the build requirements will change - it is possible that the build reqs are the main issue rather than the minimum install reqs. The build reqs for 3.7 will likely drop to 1.14 since we have moved to oldest supported NumPy.
Remove secondary check which is incompatible with using pip
and oldest-numpy-version
closes #7992
pip install git+https://github.com/statsmodels/statsmodels.gitfails withnumpy is installed but older (1.14.5) than required (1.17)#7992NumPy's guide.
Details
Notes:
needed for doc changes.
then show that it is fixed with the new code.
verify you changes are well formatted by running
flake8is installed. This command is also available on Windowsusing the Windows System for Linux once
flake8is installed in thelocal Linux environment. While passing this test is not required, it is good practice and it help
improve code quality in
statsmodels.