Landing Pages#
Audience: everyone (choose your depth)
Time: 10–30 minutes
Goal: pick the right workflow and get your first Cellucid view running
This section is intentionally written for mixed audiences:
Wet lab / non-technical users: step-by-step “do this, then that” instructions and clear success criteria.
Computational users: explicit shapes/dtypes, configuration options, performance considerations, and edge cases.
Developers/maintainers: architecture, file formats, and reproducibility notes.
Cellucid is a web app (the viewer UI) plus helper packages that bring your data into the viewer:
Cellucid web app: what you see and interact with in the browser.
cellucid(cellucid-python): this package — export/serve data and embed the web app in notebooks.cellucid-annotation: helper repo for community annotation workflows.cellucid-r: planned; not ready yet.
Choose your workflow (start here)#
“I have an AnnData and want to see it now” → Quick start (3 levels) (Level 1:
show_anndata)“I need a shareable, fast export folder (papers/collaboration)” → Quick start (3 levels) (Level 2:
prepare) + Data Preparation API (prepare/export) — The Big One“I want the viewer in a browser (no notebook)” → Installation (verify CLI) + Quick start (3 levels) (use
cellucid serve)“I need Python ↔ UI hooks (selection callbacks, programmatic highlights)” → Quick start (3 levels) (Level 3) + Jupyter Hooks System (Python ↔ Frontend)
“Something failed” → Global Troubleshooting Index (Python)
Note
If you prefer notebook-style, highly verbose walkthroughs (with lots of edge cases and troubleshooting), start with Notebooks / Tutorials (Very Detailed, Step-by-Step).
What the Python package does, how it relates to the Cellucid web app, and which workflows it supports.
How to install, optional dependencies, platform notes, and installation troubleshooting.
Copy/paste: a minimal “show”, a practical export workflow, and an advanced server + hooks workflow.
Exactly what works (and what doesn’t) across Jupyter, JupyterLab, VSCode notebooks, and Colab.