Skip to content

Instantly share code, notes, and snippets.

Embed
What would you like to do?
Generic Installation Instructions

Generic Installation Instructions

General Requirements

  • Python 3.6 or higher.
  • TensorFlow 2 if TensorFlow is used.

Instructions for macOS

The OS X Command Line Tools are required. These can be be installed with xcode-select --install. If you would like to update the Command Line Tools, delete /Library/Developer/CommandLineTools/ and run xcode-select --install.

Then install gcc and gfortran. Both are provided by brew's gcc:

brew install gcc

If you are using Anaconda, make and activate a new Anaconda environment:

conda create -n nameofenvironment python=3.8
conda activate nameofenvironment

Make sure that pip and related packages are up to date:

pip install --upgrade pip setuptools Cython numpy

If you are not using Anaconda, make and activate a new virtual environment:

pip install virtualenv
virtualenv -p python3 venv
source venv/bin/activate

Finally, install the package:

pip install package

Error: Error #15: Initializing libiomp5.dylib, but found libomp5.dylib already initialized.

I've seen this when Anaconda is used. Uninstall numpy. Then reinstall numpy as follows:

conda install nomkl
conda install numpy

Instructions for Linux

Install the Python headers. Replace in the following X with your minor version number.

sudo apt-get install python3.X-dev

Install gfortran.

sudo apt-get install gfortran

Finally, install the package.

pip install package

Instructions for Windows

The package is not tested on Windows, although success has been reported with a simple pip install package in Cygwin.

@craftablescience

This comment has been minimized.

Copy link

@craftablescience craftablescience commented Mar 31, 2021

Make sure that pip and related packages are up to date:
pip install --ugrade pip setuptools Cython numpy

Should be upgrade -- this will make it easier to copy-paste ;)

@wesselb

This comment has been minimized.

Copy link
Owner Author

@wesselb wesselb commented Mar 31, 2021

Good catch! Fixed. Thanks :)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment