Skip to content

Installation

ORCA Workbench is a Python package that installs a console command, orca-workbench. Python ≥ 3.9 with Tkinter (bundled with most Python builds) is required; the heavy dependencies (rdkit, openbabel-wheel, matplotlib) are pulled in automatically.

Pick whichever install fits.

Quick — nothing to clone

Install straight from GitHub:

pip install git+https://github.com/ACH-Repo/ACH-Orca-Workbench.git

py -m pip install git+https://github.com/ACH-Repo/ACH-Orca-Workbench.git

Then launch with orca-workbench (or python -m orca_workbench). This is the fastest path and needs no checked-out folder — but you can't edit the source in place.

Clone once, then install in editable mode so a git pull updates the app in place with no reinstall (don't move the folder afterwards — that breaks the link):

git clone https://github.com/ACH-Repo/ACH-Orca-Workbench.git
cd ACH-Orca-Workbench
module load python          # REQUIRED — see note below
pip install --user -e .
orca-workbench                 # window appears on your PC over X-forwarding

git clone https://github.com/ACH-Repo/ACH-Orca-Workbench.git
cd ACH-Orca-Workbench
py -m pip install --user -e .
py -m orca_workbench

module load python is mandatory on the cluster

The scientific stack (matplotlib → numpy) needs libopenblas.so, which that module autoloads. Without it the app fails at startup with ImportError: libopenblas.so.0: cannot open shared object file. Put module load python in your ~/.bashrc so you never forget it.

Launching on Windows

The orca-workbench command lands in a per-user Scripts folder that often isn't on PATH, so py -m orca_workbench is the reliable way to launch from anywhere. (pip prints that Scripts path during install if you'd rather add it to PATH.)

Verify the environment

orca-workbench --check-backends    # or: python -m orca_workbench --check-backends

Probes whether RDKit / OpenBabel can generate coordinates on this machine (no GUI) — run it if structure generation misbehaves. Open a saved project directly with orca-workbench myproject.json.

Slow over X-forwarding? Run the self-diagnostics

orca-workbench --diagnose launches with live timing instrumentation and writes a performance .log to your home dir when you quit — handy for pinning down why a gateway session feels sluggish.

Updating

  • Editable / cloned install: git pull in the repo folder — done. The next launch picks up the changes; no reinstall needed.

  • Quick (git+pip) install: re-run with -U:

    pip install -U git+https://github.com/ACH-Repo/ACH-Orca-Workbench.git

  • On the cluster, module load python is still required every session, whichever way you installed.