Pedagogy: add 3D rotation matrix slider (heatmap) to 40_linear_algebra_1/45_matrix_and_vector_numpy

Add ipywidgets sliders for the three Euler angles (or axis-angle parameters) of a 3D rotation matrix to `40_linear_algebra_1/45_matrix_and_vector_numpy.ipynb`. Visualize the resulting matrix as a **heatmap** (`plt.matshow` with a diverging colormap), not Hinton — this is the case where the *smooth continuous magnitudes* of `cos`/`sin` are the story, not sparsity.

## Scope

- Sliders: roll, pitch, yaw (or just one axis-angle for simplicity).
- Render: 3×3 heatmap of the rotation matrix with a diverging colormap centered at 0.
- Optionally: also render the matrix acting on a unit cube / coordinate frame in 3D for the geometric interpretation.

## Why

Per the cross-cutting visualization guideline: **heatmap for smooth continuous values; Hinton for structure/sparsity**. Rotation matrices are dense and continuous — every entry varies smoothly with the angle parameter — so this is precisely the case where heatmap is the right tool. Pairs naturally with #371 (which introduces both visualizations) by being the canonical heatmap example.

Note: this is intentionally distinct from #371 and from the matrix-operations Hinton issue. Three separate issues, three separate visual stories, all targeting the same notebook neighborhood.