A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.
Python artificial-intelligencedagdata-sciencedata-visualizationdataflowdeveloper-toolsmachine-learningnotebookspipelinepythonreactivesqlweb-app
# /// script
# requires-python = ">=3.9"
# dependencies = [
# "pymde==0.1.18",
# ]
# ///
import marimo
__generated_with = "0.19.7"
app = marimo.App()
@app.cell
def _(mo, n_items):
mo.md(
f"""
# Embedding binary trees
In this example, we'll use PyMDE to draw binary trees. Sweep the
slider to draw bigger trees ...
{n_items}
"""
)
return
@app.cell
def _(mo, n_items):
mo.md(f"Let's embed a binary tree on **{n_items.value}** vertices.")
return
@app.cell
def _(embedding, pymde, tree):
pymde.plot(embedding, edges=tree.edges)
return
@app.cell
def _(embed, make_graph, n_items):
tree, graph = make_graph(n_items.value)
embedding = embed(graph)
return embedding, tree
@app.cell
def _(mo):
n_items = mo.ui.slider(64, 1024, step=64)
return (n_items,)
@app.cell
def _(functools, pymde, torch):
@functools.cache
def make_graph(n_items):
edges = []
stack = [0]
while stack:
root = stack.pop()
c1 = root * 2 + 1
c2 = root * 2 + 2
if c1 < n_items:
edges.append((root, c1))
stack.append(c1)
if c2 < n_items:
edges.append((root, c2))
stack.append(c2)
tree = pymde.Graph.from_edges(torch.tensor(edges))
return tree, pymde.preprocess.graph.shortest_paths(tree)
return (make_graph,)
@app.cell
def _(functools, pymde):
@functools.cache
def embed(shortest_paths_graph):
mde = pymde.MDE(
shortest_paths_graph.n_items,
embedding_dim=2,
edges=shortest_paths_graph.edges,
distortion_function=pymde.losses.WeightedQuadratic(
shortest_paths_graph.distances
),
)
return mde.embed(verbose=True)
return (embed,)
@app.cell
def _():
import functools
import pymde
import torch
import marimo as mo
return functools, mo, pymde, torch
if __name__ == "__main__":
app.run()