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

# Convex combination

A weighted average of finitely many points with nonnegative weights summing to one

Let \(X\) be a real [vector space](https://knowlpedia.com/linear-algebra/vector-space/). A vector \(x\in X\) is a **convex combination** of points \(x_1,\dots,x_m\in X\) if there exist scalars \(\lambda_1,\dots,\lambda_m\ge 0\) with
\[
\lambda_1+\cdots+\lambda_m=1
\]
such that
\[
x=\sum_{i=1}^m \lambda_i x_i.
\]

**Context.** Convex combinations describe the points obtained by repeatedly taking "weighted averages." They generate the [convex hull](https://knowlpedia.com/convex-analysis/convex-hull/) of a set.

## Examples

- In \(\mathbb{R}^n\), the point \(\tfrac12x_1+\tfrac12x_2\) is the midpoint of \(x_1\) and \(x_2\).
- If \(x_1,x_2,x_3\in\mathbb{R}^2\) are vertices of a triangle, then all points in the triangle are convex combinations of \(x_1,x_2,x_3\).
