Installation Guide

Installation Guide#

TLDR

make sure you are in a virtual environment

pip install “lbm_suite2p_python[all]”

Note: The gui dependencies often install in under a minute with python 3.11 and python 3.12. On 3.10, pip will have to build wheels for imgui-bundle, which can take ~8 to 10 minutes.

For example, a gui install on Python 3.10 will build imgui-bundle from source, increasing the install time by several minutes. We recommend Python 3.11 for the greatest compatibility and install speed.

Quick Install#

LBM-Suite2p-Python has been developed to be a pure pip install.

This makes the choice of virtual-environment less relevant, you can use venv, uv (recommended), conda, it does not matter.

Tip

While this pipeline is early in development, we recommend keeping a versioon of the codebase locally using git. This will allow you to quickly pull changes and incorperate them into your environment without waiting for a pypi release.

git clone https://github.com/MillerBrainObservatory/LBM-Suite2p-Python.git
cd LBM-Suite2p-Python
pip install -e ".[all]"

pip install lbm_suite2p_python

conda create -n lbm python=3.11 -c conda-forge
conda activate lbm
git clone https://github.com/millerbrainobservatory/lbm_suite2p_python.git
cd lbm_suite2p_python
pip install -e .

GUI Dependencies#

sudo apt install libxcursor-dev libgl1-mesa-dev libglu1-mesa-dev freeglut3-dev

You will need msvcc redistributable

Troubleshooting#

Git LFS Error: smudge filter lfs failed#

If you see:

error: external filter 'git-lfs filter-process' failed
fatal: docs/source/_static/guide_hello_world.png: smudge filter lfs failed

Disable smudge during sync:

GIT_LFS_SKIP_SMUDGE=1 uv sync --all-extras --active

To debug:

git lfs logs last

This avoids downloading large binary files (e.g. images, model checkpoints) managed by Git LFS.

TODO#

  • Upload mp4 for animate traces

  • animate_traces

Contents