1import marimo
2
3__generated_with = "0.13.0"
4app = marimo.App()
5
6
7@app.cell
8def intro():
9    import marimo as mo
10    mo.md("""
11    # DAG Demo
12
13    this notebook has a non-trivial dependency graph to test DAG visualization.
14    cells form a diamond pattern: config feeds into both branch_a and branch_b,
15    which both feed into merge.
16    """)
17    return (mo,)
18
19
20@app.cell
21def config():
22    n_samples = 1000
23    seed = 42
24    return (n_samples, seed)
25
26
27@app.cell
28def branch_a(n_samples, seed):
29    import random
30    random.seed(seed)
31    values_a = [random.gauss(0, 1) for _ in range(n_samples)]
32    mean_a = sum(values_a) / len(values_a)
33    print(f"branch A: {n_samples} samples, mean={mean_a:.4f}")
34    return (values_a, mean_a)
35
36
37@app.cell
38def branch_b(n_samples, seed):
39    import random
40    random.seed(seed + 1)
41    values_b = [random.uniform(-2, 2) for _ in range(n_samples)]
42    mean_b = sum(values_b) / len(values_b)
43    print(f"branch B: {n_samples} samples, mean={mean_b:.4f}")
44    return (values_b, mean_b)
45
46
47@app.cell
48def merge(values_a, values_b, mean_a, mean_b):
49    combined = values_a + values_b
50    total_mean = sum(combined) / len(combined)
51    print(f"merged: {len(combined)} total samples")
52    print(f"  branch A mean: {mean_a:.4f}")
53    print(f"  branch B mean: {mean_b:.4f}")
54    print(f"  combined mean: {total_mean:.4f}")
55    return (combined,)
56
57
58@app.cell
59def final_stats(combined):
60    sorted_vals = sorted(combined)
61    n = len(sorted_vals)
62    median = sorted_vals[n // 2]
63    p95 = sorted_vals[int(n * 0.95)]
64    p05 = sorted_vals[int(n * 0.05)]
65    print(f"final stats:")
66    print(f"  median: {median:.4f}")
67    print(f"  p5-p95 range: [{p05:.4f}, {p95:.4f}]")
68    return ()
69
70
71if __name__ == "__main__":
72    app.run()