zsh: illegal hardware instruction python (mac m1 silicon)

[JiaqiHu2021]

I run the following code in the terminal, I get the error "zsh: illegal hardware instruction python"

Python 3.9.7 (default, Sep 16 2021, 08:50:36) 
[Clang 10.0.0 ] :: Anaconda, Inc. on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> from abess.linear import abessLogistic
>>> from sklearn.datasets import load_breast_cancer
>>> from sklearn.pipeline import Pipeline
>>> from sklearn.metrics import make_scorer, roc_auc_score
>>> from sklearn.preprocessing import PolynomialFeatures
>>> from sklearn.model_selection import GridSearchCV
>>> pipe = Pipeline([('poly', PolynomialFeatures(include_bias=False)), ('alogistic', abessLogistic())])
>>> param_grid = {'poly__interaction_only': [True, False],'poly__degree': [1, 2, 3]}
>>> scorer = make_scorer(roc_auc_score, greater_is_better=True)
>>> grid_search = GridSearchCV(pipe, param_grid, scoring=scorer, cv=5)
>>> X, y = load_breast_cancer(return_X_y=True)
>>> grid_search.fit(X, y)

[Mamba413]

Can you give me more details about your system info?

You can get them by running:

import platform
print("Platform Version: {0}, {1}".format(platform.platform(), platform.architecture()[0]))
print("Python Version:", platform.python_version())
import abess
print("Package Version:", abess.__version__)

[JiaqiHu2021]

Platform Version: macOS-10.16-x86_64-i386-64bit, 64bit
Python Version: 3.9.7
Package Version: 0.3.6

[Mamba413]

Thanks.

It is a tough issue. We have to pay some time to fix it. If you are urgent to use abess on Mac with M1 chip, you can refer to: tensorflow/tensorflow#45404, tensorflow/tensorflow#46044, or https://www.tensorflow.org/install/source#bazel_build_options, https://stackoverflow.com/questions/65770132/python3-install-of-tensorflow-on-apple-silicon-m1. Another option is using it in Windows or Linux.

[Mamba413]

The following is a step-by-step procedure for solving this issue.

1. Download miniforge from github: https://github.com/conda-forge/miniforge
   

2. Install it in terminal by commands:

mv Miniforge3-MacOSX-arm64.sh ~/
Chmod +x Miniforge3-MacOSX-arm64.sh
./Miniforge3-MacOSX-arm64.sh

3. Activate and start a new environment named abess in terminal via:

conda create --name abess python=3.8
conda activate abess

4. Follow the instruction to install the latest version of abess Python package.

After executing these steps, I believe the illegal hardware instruction python would be resolved.

If you meet any questions, feel free to contact us.

[Mamba413]

Hi @JiaqiHu2021 ，we have updated a new version in Pypi. We believe the newest version can successfully executed the following code:

>>> from abess.linear import abessLogistic
>>> from sklearn.datasets import load_breast_cancer
>>> from sklearn.pipeline import Pipeline
>>> from sklearn.metrics import make_scorer, roc_auc_score
>>> from sklearn.preprocessing import PolynomialFeatures
>>> from sklearn.model_selection import GridSearchCV
>>> pipe = Pipeline([('poly', PolynomialFeatures(include_bias=False)), ('alogistic', abessLogistic())])
>>> param_grid = {'poly__interaction_only': [True, False],'poly__degree': [1, 2, 3]}
>>> scorer = make_scorer(roc_auc_score, greater_is_better=True)
>>> grid_search = GridSearchCV(pipe, param_grid, scoring=scorer, cv=5)
>>> X, y = load_breast_cancer(return_X_y=True)
>>> grid_search.fit(X, y)

You can run

pip uninstall abess
pip install --no-cache-dir abess==0.4.2

to reinstall the latest abess library.

If you still meet any problem, feel free to contact us.

[JiaqiHu2021]

After upgrading to the latest version, I successfully ran the above code.
Thank you very much for your help!