Canonical: https://knowlpedia.com/real-analysis/lower-sum-riemann/

# Lower sum (Riemann)

A weighted sum of infima of f over subintervals of a partition.

Let \(f:[a,b]\to\mathbb{R}\) be bounded and let \(P:a=x_0<\cdots<x_n=b\) be a partition. For each subinterval, define
\[
m_i := \inf\{f(x): x\in[x_{i-1},x_i]\}.
\]

The **lower sum** of \(f\) with respect to \(P\) is
\[
L(f,P) := \sum_{i=1}^n m_i\, (x_i-x_{i-1}).
\]

[Lower sums](https://knowlpedia.com/real-analysis/lower-sum/) approximate the integral from below. As the partition is refined, lower sums increase (or stay the same).

## Examples

- If \(f(x)=x\) on \([0,1]\) and \(P:0<1/2<1\), then \(m_1=0\), \(m_2=1/2\), so \(L(f,P)=0\cdot\frac12+\frac12\cdot\frac12=\frac14\).
- If \(f(x)=c\) is constant, then \(L(f,P)=c(b-a)\) for every \(P\).
- For \(f=\mathbf{1}_{\mathbb{Q}}\) on \([0,1]\), every subinterval has \(m_i=0\), so \(L(f,P)=0\) for every \(P\).
