Closed BielStela closed 5 years ago
I created a new env without de Conda-forge channel the reproduce the old setup and everything worked fine again. This I the current env:
# packages in environment at /Users/phi/miniconda3/envs/flame_noforge:
#
# Name Version Build Channel
appdirs 1.4.3 <pip>
asn1crypto 0.24.0 py36_0
atomicwrites 1.3.0 py_0
attrs 18.2.0 py36h28b3542_0
blas 1.0 mkl
bzip2 1.0.6 h1de35cc_5
ca-certificates 2019.1.23 0
cairo 1.14.12 hc4e6be7_4
certifi 2018.11.29 py36_0
cffi 1.11.5 py36h6174b99_1
chardet 3.0.4 py36_1
cheroot 6.5.4 py36_0
cherrypy 18.1.0 py36_0
contextlib2 0.5.5 py36hd66e5e7_0
cryptography 2.5 py36ha12b0ac_0
cycler 0.10.0 py36hfc81398_0
flame 0.1 <pip>
fontconfig 2.13.0 h5d5b041_1
freetype 2.9.1 hb4e5f40_0
gettext 0.19.8.1 h15daf44_3
glib 2.56.2 hd9629dc_0
icu 58.2 h4b95b61_1
idna 2.8 py36_0
intel-openmp 2019.1 144
jaraco.functools 1.20 py36_0
jpeg 9b he5867d9_2
kiwisolver 1.0.1 py36h0a44026_0
libboost 1.65.1 hcc95346_4
libcxx 4.0.1 hcfea43d_1
libcxxabi 4.0.1 hcfea43d_1
libedit 3.1.20181209 hb402a30_0
libffi 3.2.1 h475c297_4
libgfortran 3.0.1 h93005f0_2
libiconv 1.15 hdd342a3_7
libpng 1.6.36 ha441bb4_0
libtiff 4.0.10 hcb84e12_2
libxml2 2.9.9 hab757c2_0
matplotlib 3.0.2 py36h54f8f79_0
mkl 2019.1 144
mkl_fft 1.0.10 py36h5e564d8_0
mkl_random 1.0.2 py36h27c97d8_0
more-itertools 5.0.0 py36_0
ncurses 6.1 h0a44026_1
nonconformist 1.2.5 <pip>
numpy 1.15.4 py36hacdab7b_0
numpy-base 1.15.4 py36h6575580_0
olefile 0.46 py36_0
openssl 1.1.1a h1de35cc_0
pandas 0.24.1 py36h0a44026_0
patsy 0.5.1 py36_0
pcre 8.42 h378b8a2_0
pillow 5.4.1 py36hb68e598_0
pip 19.0.1 py36_0
pixman 0.36.0 h1de35cc_0
pluggy 0.8.1 py36_0
portend 2.3 py36_0
py 1.7.0 py36_0
py-boost 1.65.1 py36h1439ea1_4
pycparser 2.19 py36_0
pyopenssl 19.0.0 py36_0
pyparsing 2.3.1 py36_0
pysocks 1.6.8 py36_0
pytest 4.2.1 py36_0
python 3.6.8 haf84260_0
python-dateutil 2.7.5 py36_0
pytz 2018.9 py36_0
pyyaml 3.13 py36h1de35cc_0
rdkit 2018.09.1.0 py36h6adbef8_1 rdkit
readline 7.0 h1de35cc_5
repoze.lru 0.7 py36_0
requests 2.21.0 py36_0
routes 2.4.1 py36_0
scikit-learn 0.20.2 py36h27c97d8_0
scipy 1.2.1 py36h1410ff5_0
setuptools 40.8.0 py36_0
simplejson 3.16.0 py36h1de35cc_0
six 1.12.0 py36_0
sqlite 3.26.0 ha441bb4_0
standardiser 0.1.7 <pip>
statsmodels 0.9.0 py36h1d22016_0
tempora 1.14 py36_0
tk 8.6.8 ha441bb4_0
tornado 5.1.1 py36h1de35cc_0
urllib3 1.24.1 py36_0
wheel 0.32.3 py36_0
xz 5.2.4 h1de35cc_4
yaml 0.1.7 hc338f04_2
zc.lockfile 1.3.0 py36h28b3542_0
zlib 1.2.11 h1de35cc_3
zstd 1.3.7 h5bba6e5_0
I will roll back the env.yml (in #117 ) to the defaults to fix the Windows issue with libfortran.
(OSX) All tests passed with default channel in env
(flame) MacBook-Air-5:flame phi$ pytest
================================================= test session starts ==================================================
platform darwin -- Python 3.7.1, pytest-4.3.0, py-1.7.0, pluggy-0.8.1
rootdir: /Users/phi/github/flame, inifile:
collected 9 items
flame/tests/test_classification.py . [ 11%]
flame/tests/test_classification_conformal.py . [ 22%]
flame/tests/test_manage.py . [ 33%]
flame/tests/test_regression.py . [ 44%]
flame/tests/test_regression_conformal.py . [ 55%]
flame/tests/test_utils.py .... [100%]
======================================== 9 passed, 2 warnings in 27.43 seconds =========================================
The environment needs to be redefined to avoid the segfault in MacOS and the dependency availability in Windows. Linked to #22