Stars2Cells now ships as a standalone application — Python and every
dependency are bundled inside. Download the file for your system, open it,
and double-click. No Conda, no pip, nothing to install.
Open the .dmg and drag Stars2Cells.app out, then double-click it.
If macOS says “unidentified developer,” right-click the app → Open once.
On an Intel Mac? Run from source below (an x86_64 build isn’t posted yet).
Unzip anywhere, open the Stars2Cells folder, and run
Stars2Cells.exe. If SmartScreen appears, click
More info → Run anyway once. The
Launch Stars2Cells.bat next to the exe adds
auto-restart-on-crash if you want it.
Which file do I pick?
Mac users: grab the .dmg — it’s the cleanest install and
keeps the app’s signature intact. The .zip is there as a fallback.
Windows users: the .zip is the only file you need — ship the whole
folder, not just the .exe.
Run from source
Prefer to run the code directly — e.g. on an Intel Mac, on Linux, or to
modify the pipeline? Clone the repository and run it in a Python environment.
Note:
The steps below use Conda because it’s the most reproducible for scientific
stacks — numpy, scipy, and faiss install as prebuilt binaries instead of
compiling from source.
Quick start (Conda)
Recommended for most source users. This creates an isolated environment so
Stars2Cells dependencies don’t affect other projects.
# 1) Clone the repository
git clone https://github.com/ariasarch/Stars2Cells_GUI.git
cd Stars2Cells_GUI
# 2) Create and activate an environment (Python 3.12 recommended)
conda create -n s2c python=3.12 -y
conda activate s2c
# 3) Install the conda-managed scientific deps first, then the rest
conda install -c conda-forge numpy=1.26.4 scipy=1.13.0 faiss-cpu -y
pip install -r requirements.txt
# 4) Launch the app
python stars2cells.py
Tip:
For auto-restart-on-crash while running from source, use the bundled launcher
instead of step 4 —
./launch_s2c.command on macOS/Linux or
launch_s2c.bat on Windows.
Quick start (pip + venv)
If you don’t use Conda, you can use Python venv — but you’ll need numpy,
scipy, and faiss-cpu available as wheels for your Python version (3.12 is the
safe choice; 3.13 has no wheels for the pinned versions and will try to compile).
# 1) Clone the repository
git clone https://github.com/ariasarch/Stars2Cells_GUI.git
cd Stars2Cells_GUI
# 2) Create and activate a virtual environment (use a Python 3.12 interpreter)
python3.12 -m venv .venv
# macOS / Linux:
source .venv/bin/activate
# Windows (PowerShell):
# .venv\Scripts\Activate.ps1
# 3) Install dependencies
pip install numpy==1.26.4 scipy==1.13.0 faiss-cpu
pip install -r requirements.txt
# 4) Launch the app
python stars2cells.py
Recommended environment
Python 3.12 is recommended (3.10–3.12 supported). Stars2Cells depends on
common scientific Python packages plus PyQt5 for the interface. The packaged
downloads above already contain all of this — the source route is only for
development or unsupported platforms.
