Skip to content

Getting started

This document provides a detailed guide for installing PickET.

Dependencies

  • Python dependencies can be found in pyproject.toml.
  • TomoEED

Instructions for installing the Python dependencies, CuPy and TomoEED are provided below:

Installation

1. Create a new virtual environment

In a new terminal window, run the following command to install venv package: 

pip install venv

Note

On Ubuntu-like operating systems, this command needs to be run as shown here: 

sudo apt install python3-venv

Then, run the following commands sequentially to create a new virtual environment: 

mkdir <path_to_new_virtual_environment>
cd <path_to_new_virtual_environment>
python -m venv picket_env

2. Activate the environment 

source <path_to_new_virtual_environment>/bin/activate

Note

Make sure to activate the environment before installing rest of the dependencies and before running PickET. The virtual environment can be deactivated by running deactivate in the terminal or simply by closing the terminal window.

3. Install CuPy

If the CUDA version is 12.<something>, install CuPy by running

pip install cupy-cuda12x

Else if CUDA version is 11.<something>, install CuPy by running.

pip install cupy-cuda11x

Note

The CUDA version can be checked by running nvidia-smi from the terminal. It will be shown on the top right corner in the generated output.

4. Install all the PickET

git clone https://github.com/isblab/pickET.git
cd pickET
pip install .
Installing PickET via pip using the above command builds executables for the two steps in a PickET run - generating semantic segmentation and localizing particles. (See also Usage instructions)

5. Install TomoEED

  1. Visit the official TomoEED webpage.

  2. Fill and submit the Google Form.

  3. Copy the generated download link and paste it in a new browser tab. This will download TomoEED as a zipped directory. Unzip it.

Back to Home
Go to usage instructions