Canonical: https://knowlpedia.com/convex-analysis/seminorm/

# Seminorm

A subadditive, absolutely homogeneous function p(λx)=|λ|p(x).

Let \(X\) be a [vector space](https://knowlpedia.com/linear-algebra/vector-space/) over \(\mathbb{R}\) or \(\mathbb{C}\). A function \(p:X\to\mathbb{R}\) is called a **seminorm** if:

1. (**Subadditivity**) \(p(x+y)\le p(x)+p(y)\) for all \(x,y\in X\).
2. (**Absolute homogeneity**) \(p(\lambda x)=|\lambda|\,p(x)\) for all \(x\in X\) and all scalars \(\lambda\).

Every [norm](https://knowlpedia.com/convex-analysis/norm-normed-vector-space/) is a seminorm, but a seminorm may vanish on nonzero vectors (e.g., \(p(x_1,x_2)=|x_1|\) on \(\mathbb{R}^2\)).

Seminorms are the natural domination bounds in the seminorm versions of Hahn–Banach, including [the real seminorm Hahn–Banach theorem](https://knowlpedia.com/convex-analysis/hahn-banach-extension-dominated-by-a-seminorm-real-case/) and [the complex seminorm Hahn–Banach theorem](https://knowlpedia.com/convex-analysis/hahn-banach-theorem-in-complex-vector-spaces/).
