Closed simonprovost closed 1 year ago
Hi @simonprovost, please consider to lower the version of Cython and see if it helps.
According to the release history of Cython, the version number directly increases from 0.29
to 3.0
, and there could be a lot of compatiability problems.
Hello @xuyxu,
Thank you for your swift reply 👌 I attempted to set up a conda environment using Python 3.9.18 and tried installing Cython versions 0.20 through 0.28 using the command python -m pip install cython==0.X
, where X
∈ {20, 21, 22, ..., 28}
. However, all attempts resulted in unsuccessful installations accompanied by numerous errors.
I would greatly appreciate it if you could share the specific versions of Python and Cython that are compatible with DeepForest. This information will help us troubleshoot further, and we'll be able to provide logs if we encounter any additional issues.
Cheers 👋
Maybe provide your conda env pip list, we'll try to mirror that?
I have tested on my Macbook Air (M2), and the classification demo in README.md
works fine.
IMPORTANT: There is no need to build the library from source, please use pip install deep-forest
.
Fantastic! We're making progress 👌
I've successfully installed numpy, scikit, joblib, and your specified version of Cython. However, I hit a snag when pip install deep-forest
failed. This was the initial issue I faced before I attempted solution #6.
Here are additional details, given that I am operating within a conda environment activated under Python 3.9:
/pip list (TMP_ENV)
Package Version
------------------ ---------
Brotli 1.1.0
certifi 2023.7.22
charset-normalizer 3.3.0
Cython 0.29.28
idna 3.4
joblib 1.1.0
numpy 1.21.5
packaging 23.2
pip 23.2.1
platformdirs 3.11.0
pooch 1.7.0
PySocks 1.7.1
requests 2.31.0
scikit-learn 1.0.2
scipy 1.10.1
setuptools 68.2.2
threadpoolctl 3.2.0
typing_extensions 4.8.0
urllib3 2.0.6
wheel 0.41.2
pip --version (TMP_ENV)
pip 23.2.1 from [...]miniforge3/envs/TMP_ENV/lib/python3.9/site-packages/pip (python 3.9)
pip install deep-forest (TMP_ENV)
ERROR: Could not find a version that satisfies the requirement deep-forest (from versions: none)
# Note: I attempted specifying deep-forest==0.1.7 and others, yet no luck—it seems unable to locate the actual pip package.
pip install deep-forest==0.7 (TMP_ENV)
ERROR: Could not find a version that satisfies the requirement deep-forest==0.1.7 (from versions: none)
ERROR: No matching distribution found for deep-forest==0.1.7
Any insights? We seem to be on the brink of resolving this !
I greatly appreciate your assistance once more @xuyxu!
Cheers,
Here is the running result of mine:
Could you download this wheel from here, and see if pip install {name}.wheel
works.
I am eagerly looking forward to having deep-forest
operational on my machine as well but unfortunately have again encountered some hurdles during the installation process. Despite successfully resolving the prior Cython issue, I am now facing challenges with installing the deep-forest
wheel. See below further👌
OSX: 14.0 (23A344)
CPU: Apple M2 (8 cores, 8 threads)
conda version: 23.7.2
pip version: 23.2.1 (with PY version: Python 3.9)
I've tried both curl and pip commands to install the proposed wheel, but to no avail. Here's the error message I received:
ERROR: deep_forest-0.1.7-cp39-cp39-macosx_10_9_x86_64.whl is not a supported wheel on this platform.
Despite the error, I also attempted to install other available wheels, although they were not aligned with my Python version (3.9) and OS but just by any curiosity. Yet to no success again. Lastly, I can also confirm that my pip and wheel are updated, yet the issue persists.
Any insights or suggestions.. am afraid that you come to a dead end at some points, but thanks in any event to your great support !
EDIT 1.0: I just realised that I am using miniforge3 instead of anaconda3 package manager. Might that be the case? I am far from convinced though
Could you try if using the Python and pip from Anaconda (which provides a version optimized for M1/M2) helps? You can find the download link here.
Hi @xuyxu ,
Downloaded Anaconda and ensured my newly created conda environment was contained within anaconda and not miniforge anymore. I have installed the prerequisites you provided above. Then, I attempted to install, by ensuring that the pip of the conda environment is functioning properly prior to anything (i.e., making sure this is the pip of the anaconda env), deep-forest, yet I am having the same error:
[...]/anaconda3/envs/deepforesrt/bin/pip install deep-forest $
ERROR: Could not find a version that satisfies the requirement deep-forest (from versions: none)
ERROR: No matching distribution found for deep-forest
Next, I cloned the repository and attempted to compile by source using the current anaconda environment, but the same error occurred as per the cython splitter based issue mentionned earlier.
What is wrong with my situation 🧐 I am very much lost. Soon, I'll attempt to reinstall all that on a university-owned UNIX cloud workstation, but I encountered the same error as described in this message a few days ago.
If you have last thought, I'd be so very much grateful :)
So sorry for this annoying issue ! Cheers
ANACONDA V (General Conda Info):
conda info (deepforesrt)
active environment : deepforesrt
active env location : [...]/anaconda3/envs/deepforesrt
shell level : 3
user config file : [...]/.condarc
populated config files : [...]/.condarc
conda version : 23.7.4
conda-build version : 3.26.1
python version : 3.11.5.final.0
virtual packages : __archspec=1=arm64
__osx=14.0=0
__unix=0=0
base environment : [...]/anaconda3 (writable)
conda av data dir : [...]/anaconda3/etc/conda
conda av metadata url : None
channel URLs : https://conda.anaconda.org/conda-forge/osx-arm64
https://conda.anaconda.org/conda-forge/noarch
package cache : [...]/anaconda3/pkgs
[...]/.conda/pkgs
envs directories :[...]/anaconda3/envs
[...]/.conda/envs
platform : osx-arm64
user-agent : conda/23.7.4 requests/2.31.0 CPython/3.11.5 Darwin/23.0.0 OSX/14.0 aau/0.4.2 c/EvJX_zyd1e6meTODjVe_qw s/lBpQahJvyK8y8vpo6U_X8w e/pWa2DGIJipu6UJWQfGB__Q
UID:GID : 502:20
netrc file : None
offline mode : False
ANACONDA PIP V:
[..]/anaconda3/envs/deepforesrt/bin/pip --version (deepforesrt)
pip 23.2.1 from [..]/anaconda3/envs/deepforesrt/lib/python3.9/site-packages/pip (python 3.9)
Here is my conda info
:
(base) xuyxu@Yi-XuandeAir ~ % conda info
active environment : base
active env location : /Users/xuyxu/opt/anaconda3
shell level : 1
user config file : /Users/xuyxu/.condarc
populated config files :
conda version : 4.13.0
conda-build version : 3.21.8
python version : 3.9.12.final.0
virtual packages : __osx=10.16=0
__unix=0=0
__archspec=1=x86_64
base environment : /Users/xuyxu/opt/anaconda3 (writable)
conda av data dir : /Users/xuyxu/opt/anaconda3/etc/conda
conda av metadata url : None
channel URLs : https://repo.anaconda.com/pkgs/main/osx-64
https://repo.anaconda.com/pkgs/main/noarch
https://repo.anaconda.com/pkgs/r/osx-64
https://repo.anaconda.com/pkgs/r/noarch
package cache : /Users/xuyxu/opt/anaconda3/pkgs
/Users/xuyxu/.conda/pkgs
envs directories : /Users/xuyxu/opt/anaconda3/envs
/Users/xuyxu/.conda/envs
platform : osx-64
user-agent : conda/4.13.0 requests/2.27.1 CPython/3.9.12 Darwin/23.0.0 OSX/10.16
UID:GID : 501:20
netrc file : None
offline mode : False
and here is my pip version
:
(base) xuyxu@Yi-XuandeAir ~ % pip --version
pip 21.2.4 from /Users/xuyxu/opt/anaconda3/lib/python3.9/site-packages/pip (python 3.9)
It looks like your virtual environment is with Python 3.11, which is too new. Also, it lacks the x86 channel for searching packages. Maybe you should install the Anaconda with Intel Chip ? (see the difference on tag platform
between us)
Hello @xuyxu
I am appreciative of your support !! things ultimately went smoothly. I am able to run DeepForest :)
For readers experiencing the same issue. I would suggest uninstalling miniforge3 and anaconda from your computer. Download anaconda again, Intel or Apple silicon (not tested), and repeat the procedure. Make sure to try matching as much as possible @xuyxu's configurations and you should be fine (especially python 3.9 and Cython 0.29.X).
@xuyxu any perspective for adapting DF to Python 3.10 based projects?
Cheers
Dear Authors,
Issue:
I've been facing persistent installation issues with the Deep Forest package on my new Macbook Pro 2022 (Apple Silicon, M2, 13-inch) running the latest macOS Sonoma and on a stable Ubuntu version (cluster university version no physical machine like a laptop or PC). This is despite having successfully installed and used the library on an older non-Apple Silicon MacBook a year or two ago..
Steps Tried:
pip
.Error Encountered:
The consistent error I encountered points to
splitter.pyx
of the Cython side. Here's the error for reference:Additional Information:
I had previously managed to get the library working on an older MacBook (non-Apple Silicon) a year or two ago without any issues. However, I've spent a considerable amount of time today trying to get it working on my current system, and I am quite frustrated at the lack of progress.
Would appreciate very much any guidance or potential solutions to address this issue.
Thank you so much for your great novelty with Deep forest :)
Cheers