Canonical: https://knowlpedia.com/convex-analysis/direct-sum-of-subspaces/

# Direct sum of subspaces

A sum of subspaces with trivial intersection

Let \(X\) be a vector space and let \(M,N\) be [linear subspaces](https://knowlpedia.com/convex-analysis/linear-subspace/).

Let \(Y:=M+N\) be their (set) sum. We say that \(Y\) is the **direct sum** of \(M\) and \(N\) if
\[
M\cap N=\{0\}.
\]
In this case we write
\[
Y=M\oplus N.
\]

Direct sums formalize "adding independent directions": if \(M\cap N=\{0\}\), then no nonzero vector lies in both subspaces. The key structural property is uniqueness of decomposition, captured in [the direct sum characterization](https://knowlpedia.com/convex-analysis/characterization-of-direct-sums/).

## Examples

- In \(\mathbb{R}^2\), let \(M=\operatorname{span}\{(1,0)\}\) and \(N=\operatorname{span}\{(0,1)\}\). Then \(\mathbb{R}^2=M\oplus N\).
- In \(\mathbb{R}^3\), the \(xy\)-plane and the \(z\)-axis form a direct sum: \(\mathbb{R}^3=\mathbb{R}^2\times\{0\}\ \oplus\ \operatorname{span}\{(0,0,1)\}\).
- If \(M=N\neq\{0\}\), then \(M+N=M\) but \(M\cap N=M\neq\{0\}\), so this is not a direct sum.
