Skip to content

Latest commit

 

History

History
149 lines (96 loc) · 5.87 KB

README.md

File metadata and controls

149 lines (96 loc) · 5.87 KB

Install instructions of anaconda environment

Most of the libraries related to the OpenBCI hardware works best on Windows machines. Some of the required dependencies aren't even available on macOS, for this reason these install instructions focus mainly on Windows 10 machines. However, a conda environment for macOS is also made available as well as some notes for macOS specific installs.

Table of contents

Author

Name Role VUB mail Personal mail
Lennert Bontinck Master Thesis writer [email protected] [email protected]

OpenBCI GUI

Installing OpenBCI GUI should be straightforward and the steps can be followed from the online doc.

  • Install CYTON DONGLE DRIVERSfrom the OpenBCI website or from the FTDI Chip website. The FTDI chip website has an easy setup executable.
    • Version 2.12.36.4 was used.
    • MacOS note: place the installer in the /application folder before opening it. The macOS installer is in Beta as of September 2021 and requires this hack to work correctly.
  • Instal OpenBCI GUI.
    • Version 5.0.8 was used.
    • MacOS note: V5.0.6, which was the latest version as of September 2021, has a bug on macOS where only a white screen is shown. Using the older V5.0.5 should work. This older version can be found under the GitHub releases.

Cuda drivers

Having the Nvidia Cuda drivers installed can help increase the speed of the Deel Learning (DL) models. Installing Cuda depends on the system used and thus no exact instructions are given. Check the TensorFlow doc for more information.

For our Windows 10 machine with an NVidia GTX970 the installation instructions from Aladdin Persson on YouTube were followed. The Cuda version used is 11.6, which was the latest at the moment, but the CuDNN version used was 8.2.2 (July release), as other versions seem to give errors.

MacOS specific preruiqisite

To use the PyQt5 backend for Matplotlib, which is recommended for MNE plotting, pyqt needs to be installed via Homebrew.

brew install pyqt

Setting up Anaconda environment

The instructions below highlight the steps needed to recreate the used anaconda environment.

Configuring the base environment

  • Install the free version of Anaconda Navigator. V2.1.4 was used.

  • From the Anaconda Navigator GUI, create a new environment named bci-master-thesis.

    • Include both Python and R. The following versions were used:
      • Python 3.8.10
      • R 3.6.1
    • Doing so should install a whole suite of packages by default

  • Using the Anaconda Prompt (Anaconda3) application, activate the newly created environment.

    • NOTE: it might be required to run the prompt as administrator for all of the below steps.

    • # Activates the previously created bci-master-thesis Anaconda environment.
conda activate bci-master-thesis

  • Install some conda available packages on the environment

    • # Pandas is a famous Python Data Analysis Library and is used by a lot of other packages.
# The following command installs Pandas and its dependencies. V1.4.1 was used.
conda install pandas=1.4.1

# We install pip to install packages not available from conda install. V21.2.2 was used.
conda install pip=21.2.2

  • Install some pip available packages on the environment

    • # MNE is a famous Python package for visualising and working with neurophysiological data.
# V1.0.2 was used.
pip install mne==1.0.2

# Install Matplotlib for plotting purposes.
# V3.5.1 was used.
pip install matplotlib==3.5.1

# Install backend for matplotlib visualisations.
# V5.15.6 was used.
pip install PyQt5==5.15.6

# Install Sci-Kit learn for ML.
# V1.0.2 was used.
pip install scikit-learn==1.0.2

# Install LightGBM for ML.
# V3.3.2 was used.
# NOTE: on mac you might need to 'brew install libomp' to get lightgbm to work
pip install lightgbm==3.3.2

# Install Pyriemann for EEG signal processing.
# V0.2.7 was used.
pip install pyriemann==0.2.7

# Install TensorFlow for Deep Learning
# V2.8.0 was used.
pip install tensorflow==2.8.0

Configuring the environment from the YML file

The anaconda Windows environment is also exported to the bci-master-thesis-environment-windows.yml YML file. This file is available here. You can load it in via the terminal as follows:

# Navigate to the folder where the YML file is located
cd bci-master-thesis/documentation/installation/environments
# Configure a new environment from the YML file
# These were exported using: conda env export > file.yml --no-builds
## Note: macOS variant of file available
conda env create -f bci-master-thesis-environment-windows.yml

© Lennert Bontinck VUB 2021-2022