Setting Up the Environment¶

This section describes how to prepare your environment for running prismPYP.

1. Install Conda¶

Download and install Miniconda (recommended) or Anaconda.

Verify your installation:

conda --version

2. Clone the prismPYP Repository¶

git clone https://github.com/nextpyp/prismpyp.git
cd prismpyp

3. Create and Activate the prismPYP Environment¶

Using Conda and pip:

conda create -n prismpyp -c conda-forge python=3.12 pip
conda activate prismpyp

4. Install prismPYP and Dependencies¶

Install the core package and dependencies with GPU support:

python -m pip install -e . --extra-index-url https://download.pytorch.org/whl/cu121

Note

This ensures PyTorch is installed with the correct CUDA 12.1 build.

Then install FAISS-GPU (via Conda, since pip wheels for Python 3.12 are unsupported):

conda install -c pytorch -c conda-forge faiss-gpu=1.9.0

5. Create and Install the Phoenix Environment¶

Phoenix enables interactive 3D visualization and manual selection of micrographs directly within the embedding space.

For the best performance, Phoenix should be installed and run locally (not on a remote cluster).

Info

Phoenix provides an intuitive interface to explore embeddings, filter high-quality micrographs, and export subsets for further refinement.

Install Phoenix using pip (make sure Conda and pip are installed in your local environment).

conda create -n phoenix -c conda-forge python=3.8 pip
conda activate phoenix
wget https://github.com/nextpyp/prismpyp/blob/main/requirements-phoenix.txt -O requirements-phoenix.txt
python -m pip install -r requirements-phoenix.txt

Your environments are now ready for use!

Continue to the next step to learn how to prepare and organize input data for prismPYP.