When installing OpenWillis, you have the option to install support for all modalities (video, audio, GPS, etc.). This is covered in Option A. However, most users don’t need this, and it can make the library unnecessarily heavy. So, we created Option B, which allows you to install sub-packages specific to what you need to use the library for (e.g., just for audio processing).
<aside> ℹ️
Note
Make sure python and anaconda are installed on your machine before continuing. This resource provides detailed instructions on installing python; this resource provides detailed instructions for installing anaconda.
</aside>
We highly recommend a dedicated virtual environment for OpenWillis. So, we’ll start by creating this environment and installing the necessary dependencies.
Note: Throughout installation, if prompted to ‘Proceed?’ type ‘y’ and ‘return’ or ‘enter’.
Step 1
Create and enter a virtual environment for OpenWillis. In a new terminal window, run the following:
conda create -n openwillis_env python=3.10
# Note: the full version of OpenWillis requires python version 3.9 or 3.10 even though more recent versions of python are available
conda activate openwillis_env
Step 2
After creating and activating the virtual environment in Step 1, proceed with installing the required packages ffmpeg, portaudio, pysoundfile, and forest:
conda install -c conda-forge ffmpeg
conda install portaudio
conda install pysoundfile -c conda-forge
conda install git pip
pip install git+https://github.com/onnela-lab/forest@5b9ba664cc09ac49f58e42cce3b271aa03a41967
Step 3
Now, install OpenWillis:
pip install openwillis
Step 3 - part 2, for Linux users only
Linux users will also need to run the following:
sudo apt-get update
sudo apt-get install libgl1-mesa-glx
Step 4