Canonical: https://knowlpedia.com/real-analysis/total-variation/

# Total variation

The supremum of sums of absolute increments over all partitions.

The **total variation** of a function \(\alpha:[a,b]\to\mathbb R\) is the extended nonnegative number
\[
V_a^b(\alpha)=\sup_P \sum_{i=1}^n |\alpha(x_i)-\alpha(x_{i-1})|,
\]
where the supremum is taken over all [partitions](https://knowlpedia.com/real-analysis/partition-of-an-interval/) \(P=\{a=x_0<\cdots<x_n=b\}\).

## Remarks

Total variation measures how much \(\alpha\) accumulates change along the interval and is the defining quantity for a [function of bounded variation](https://knowlpedia.com/real-analysis/bounded-variation-function/). For [monotone](https://knowlpedia.com/real-analysis/monotone-function/) functions it reduces to the net change.

## Examples

- If \(\alpha\) is increasing on \([a,b]\), then \(V_a^b(\alpha)=\alpha(b)-\alpha(a)\).
- For \(\alpha(x)=\sin x\) on \([0,2\pi]\), one has \(V_0^{2\pi}(\alpha)=4\).
