-
Notifications
You must be signed in to change notification settings - Fork 123
Installation fails on python3.7 #204
Comments
hey, i just solved this problem. |
or just clone the repo and replace see issue 191 |
@sparkdoc the clone option does not work for me (Python 3.7.9, GCC 8.3.0, cython 0.29.24). The cythonize flag still returns the same traceback as trying to install from PyPI
|
@pzivich
|
@kevin-dietz it looks like it installs (ignoring the warnings for unitialized vars or decprecated numpy API). However, when I try to import, I get the following error >>> from pyearth import Earth
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/mnt/shared/code/PythonLibs/py-earth/pyearth/__init__.py", line 6, in <module>
from .earth import Earth
File "/mnt/shared/code/PythonLibs/py-earth/pyearth/earth.py", line 1, in <module>
from ._forward import ForwardPasser
ModuleNotFoundError: No module named 'pyearth._forward' |
@pzivich One of two issues come to mind. You may have to manually move the pyearth directory that the install creates (e.g. Or you could have installed it into the wrong python version. You can use the actual python path to be explicit -- e.g. |
Regarding the below issue of "No module named pyearth._forward" ... I found a solution in stackoverflow somewhere. When you create the example script to test the py-earth build (as in the readme), do not put that example script in the py-earth folder, put it in your home dir or some other dir. Then run it from there.
|
Dear @jcrudy
I tried several ways to install py-earth but all failed. I have Win10 and a python 3.7 environment.
Firstly, I used the git clone repository method to install, it gives me: Microsoft Visual C++ 14.0 is required. Get it with "Build Tools for Visual Studio". I don't want to install visual studio because it's too big (7GB)
Secondly, I used "conda install --channel "conda-forge" sklearn-contrib-py-earth" to install, it gives me: "An HTTP error occurred when trying to retrieve this URL."
Finally, I used I used the "sklearn_contrib_py_earth-0.1.0-cp36-cp36m-in_amd64" to install, it gives me: "ERROR: sklearn_contrib_py_earth-0.1.0-cp36-cp36m-win_amd64.whl is not a supported wheel on this platform."
I am using anaconda, and can you give me guidance on py-earth installation on python 3.7 environments?
Many Thanks
The text was updated successfully, but these errors were encountered: