Canonical: https://knowlpedia.com/linear-algebra/hadamard-product/

# Hadamard product

Entrywise multiplication of vectors or matrices of the same shape.

For two [matrices](https://knowlpedia.com/linear-algebra/matrix/) \(A,B\) of the same shape, the **Hadamard product** is their entrywise product
\[
(A\odot B)_{ij}=A_{ij}B_{ij}.
\]
For vectors, \((a\odot b)_i=a_i b_i\). This is a different operation from ordinary matrix multiplication, which sums over an intermediate index.

## Coordinate dependence

Entrywise multiplication uses the chosen coordinates. For a fixed vector \(a\), the map \(b\mapsto a\odot b\) is multiplication by the diagonal matrix with entries \(a_i\). Thus differentiating coordinatewise squares gives \(D(y\mapsto y\odot y)_y[h]=2y\odot h\) over the real or complex numbers.

## References

- [Sheldon Axler, Linear Algebra Done Right, 4th ed., Chapters 2–3, 6 and 9](https://linear.axler.net/LADR4e.pdf).
