Marimo

What is Marimo?

Marimo is a modern, reactive Python notebook.

Think of it as Jupyter’s more reproducible, maintainable cousin:

• Cells automatically re-execute if their dependencies change.
• Everything lives in a plain .py script — version control friendly.
• You run it locally; view it in your browser.
• Easy to share: .py → .html → send to anyone.

It’s great for:

• Data analysis
• Quick experiments
• Dashboard-like reports
• Reproducible research

✅ Step-by-Step: Make your own env the default for Marimo

Activate your environment

For .venv:

# From your project root
source .venv/bin/activate

For Conda:

conda activate myenv  # replace 'myenv' with your env name

Install ipykernel inside your env

This makes your environment show up as a kernel option in Marimo (and Jupyter):

pip install ipykernel

Register your environment as a new kernel

Still inside your activated env, run:

python -m ipykernel install --user --name=.venv --display-name "Python (.venv)"

• -name=myenv: internal name — can be anything, no spaces.
• -display-name: friendly name you’ll see in Marimo → use something like Python (myenv) or Python (.venv).

✅ Now your kernel is installed!

Open Marimo and switch the kernel

• Open your .marimo notebook.
• Go to the Kernel or Interpreter menu (top bar).
• Pick Python (myenv) or whatever display name you chose.
• Restart the kernel if asked.

5️⃣ Confirm it’s using YOUR env

Run this in a cell:

python
CopyEdit
import sys
print(sys.executable)

You should see a path like:

swift
CopyEdit
/path/to/your/project/.venv/bin/python

or

swift
CopyEdit
/Users/you/miniconda3/envs/myenv/bin/python

If you still see something like /usr/bin/python or /path/to/marimo, you’re still on the default kernel → switch it again.

6️⃣ (Optional) Disable Conda base if you use Conda

To avoid base conflicting with your custom env, run once:

bash
CopyEdit
conda config --set auto_activate_base false

✅ Quick checklist

✔️ Activate env

✔️ Install ipykernel

✔️ Register kernel with python -m ipykernel install

✔️ Switch kernel in Marimo

✔️ Confirm with sys.executable

🔥 Result

Now, all your Marimo notebooks run inside your real environment — no more surprises, no more duplicated installs.

This is the best practice for real projects, version control, and reproducibility.

If you want, I can help you write a small Makefile or shell alias to automate this setup — just say so!

Ask ChatGPT

✅ 🔑 Marimo Commands Cheat Sheet

Marimo’s CLI is very clean — here are the main commands, when to use them, and extra flags you should know.

🗂️ 1️⃣ marimo run

What it does:

Launches your .py notebook as a local interactive server in your browser.

Basic use:

bash
CopyEdit
marimo run my_notebook.py

Useful flags:

✏️ 2️⃣ marimo edit

What it does:

Opens an existing Marimo notebook in your default editor.

Use case:

This just opens the .py file in whatever editor is set as default for .py — so honestly, it’s optional. Many people just use VSCode, PyCharm, or vim directly.

Example:

bash
CopyEdit
marimo edit my_notebook.py

📦 3️⃣ marimo export

What it does:

Converts your .py notebook to a static HTML so you can share it with anyone.

Basic use:

bash
CopyEdit
marimo export my_notebook.py my_notebook.html

Pro tip:

Before exporting, make sure your notebook runs cleanly so that outputs are saved in the HTML.

🚀 4️⃣ marimo new

What it does:

Creates a new starter notebook with example cells.

Example:

bash
CopyEdit
marimo new my_new_notebook.py

Tip:

This adds a few example cells to help you get started. You can delete or modify them.

🧩 5️⃣ marimo --help

What it does:

Shows you all available commands and options in your version.

Example:

bash
CopyEdit
marimo --help

✅ Summary Table: All Marimo CLI Commands

✅ Tips for real usage

🔗 Most used:

marimo run and marimo export — these 2 cover 90% of your workflow.

⚡ Ports:

If you run multiple notebooks, always specify different ports:

bash
CopyEdit
marimo run notebook1.py --port 2718
marimo run notebook2.py --port 3000

💡 Use your own editor:

marimo edit is optional — it’s faster to just open .py files in VSCode, Sublime, or PyCharm.

🌐 Sharing on LAN:

Use --host 0.0.0.0 + firewall open → share your notebook with others on your Wi-Fi:

bash
CopyEdit
marimo run my.py --host 0.0.0.0 --port 8000

Then visit http://<your-local-ip>:8000 from another device.

✅ Add to your Marimo Guide

🗂️ Core CLI: run, edit, export, new, --help

🚀 Always remember: run for interactive, export for static sharing, new for quick starters.