Windows
The first requirement is to install Python 3.8 or 3.9. Our recommendation is to download and install Python from the official website, with the following steps.
Python installation
Step 1: Download Python
To start, go to python.org/downloads and then click on the button to download the version of Python you need (3.8 or 3.9).
Step 2: Run the .exe file
Next, run the .exe file that you just downloaded, and then follow the installation instructions.
Note that you should also check the box to add Python to the Path, and make sure that pip is also installed (you can check this by running pip help). If not, use the below command to install:
python3 -m ensurepip --upgradeBitfount Installation
Once Python is installed, you should be able to install bitfount from pypi.
Our recommendation is to install and use Bitfount in a virtual environment, see instructions on how to do this here.
Now, use pip to install bitfount.
pip install bitfountIf you are planning on using the bitfount package with Jupyter Notebooks we recommend you install extra tutorials dependencies via bitfount[tutorials] which will make sure you are running compatible Jupyter dependencies.
pip install 'bitfount[tutorials]'If you want to use differential privacy (DP), you will need to install the DP extras as well:
pip install 'bitfount[dp]'
Other libraries
torchcsprng
The torchcsprng library is required by bitfount for Cryptographically Secure Pseudo-Random Number Generation. On Windows, the installation of this library needs to be done slightly differently depending on whether you are on CPU or to match your CUDA version if present. If you have any issues, please follow the instructions here on how best to install the library for your environment.
Next Steps
Once you’ve installed Bitfount, whoever will provide the data for analysis must connect the data to a Pod. See For Data Custodians for detailed instructions.
If data is already connected to a Pod for analysis and you wish to train models on or query it, see For Data Scientists.