# Definition:Summation

## Contents

## Definition

Let $\struct {S, +}$ be an algebraic structure where the operation $+$ is an operation derived from, or arising from, the addition operation on the natural numbers.

Let $\tuple {a_1, a_2, \ldots, a_n} \in S^n$ be an ordered $n$-tuple in $S$.

### Definition by Index

The composite is called the **summation** of $\tuple {a_1, a_2, \ldots, a_n}$, and is written:

- $\displaystyle \sum_{j \mathop = 1}^n a_j = \paren {a_1 + a_2 + \cdots + a_n}$

### Definition by Inequality

The **summation** of $\left({a_1, a_2, \ldots, a_n}\right)$ can be written:

- $\displaystyle \sum_{1 \mathop \le j \mathop \le n} a_j = \left({a_1 + a_2 + \cdots + a_n}\right)$

### Definition by Propositional Function

Let $R \left({j}\right)$ be a propositional function of $j$.

Then we can write the **summation** as:

- $\displaystyle \sum_{R \left({j}\right)} a_j = \text{ The sum of all $a_j$ such that $R \left({j}\right)$ holds}$.

If more than one propositional function is written under the summation sign, they must *all* hold.

## Infinite

Let an infinite number of values of $j$ satisfy the propositional function $R \left({j}\right)$.

Then the precise meaning of $\displaystyle \sum_{R \left({j}\right)} a_j$ is:

- $\displaystyle \sum_{R \left({j}\right)} a_j = \left({\lim_{n \mathop \to \infty} \sum_{\substack {R \left({j}\right) \\ -n \mathop \le j \mathop < 0}} a_j}\right) + \left({\lim_{n \mathop \to \infty} \sum_{\substack {R \left({j}\right) \\ 0 \mathop \le j \mathop \le n} } a_j}\right)$

provided that both limits exist.

If either limit *does* fail to exist, then the **infinite summation** does not exist.

## Index Variable

Consider the **summation**, in either of the three forms:

- $\displaystyle \sum_{j \mathop = 1}^n a_j \qquad \sum_{1 \mathop \le j \mathop \le n} a_j \qquad \sum_{\map R j} a_j$

The variable $j$, an example of a bound variable, is known as the **index variable** of the summation.

## Summand

The set of elements $\set {a_j \in S: 1 \le j \le n, \map R j}$ is called the **summand**.

## Vacuous Summation

Take the summation:

- $\displaystyle \sum _{\Phi \left({j}\right)} a_j$

where $\Phi \left({j}\right)$ is a propositional function of $j$.

Suppose that there are no values of $j$ for which $\Phi \left({j}\right)$ is true.

Then $\displaystyle \sum_{\Phi \left({j}\right)} a_j$ is defined as being $0$.

This summation is called a **vacuous summation**.

This is because:

- $\forall a: a + 0 = a$

where $a$ is a number.

Hence for all $j$ for which $\Phi \left({j}\right)$ is false, the sum is unaffected.

This is most frequently seen in the form:

- $\displaystyle \sum_{j \mathop = m}^n a_j = 0$

where $m > n$.

In this case, $j$ can not at the same time be both greater than or equal to $m$ and less than or equal to $n$.

Some sources consider such a treatment as abuse of notation.

## Also known as

It is common for the term **sum** to be used for **summation**.

However, despite the fact that **summation** is longer, and a less-aesthetic neologism for **sum**, the term **summation** is preferred on $\mathsf{Pr} \infty \mathsf{fWiki}$ for immediate clarity.

## Notation

The sign $\sum$ is called **the summation sign** and sometimes referred to as **sigma** (as that is its name in Greek).

## Also see

- Definition:Product Notation (Algebra)
- Definition:Composite (Abstract Algebra)
- Definition:Series
- Results about
**summations**can be found here.

## Historical Note

The notation $\sum$ for a summation was famously introduced by Joseph Fourier in $1820$:

*Le signe $\displaystyle \sum_{i \mathop = 1}^{i \mathop = \infty}$ indique que l'on doit donner au nombre entier $i$ toutes les valeurs $1, 2, 3, \ldots$, et prendre la somme des termes.*

- (
*The sign $\displaystyle \sum_{i \mathop = 1}^{i \mathop = \infty}$ indicates that one must give to the whole number $i$ all the values $1, 2, 3, \ldots$, and take the sum of the terms.*)

- -- 1820:
*Refroidissement séculaire du globe terrestre*(*Bulletin des Sciences par la Société Philomathique de Paris***Vol. 3, 7**: pp. 58 – 70)

- -- 1820:

However, some sources suggest that it was in fact first introduced by Euler.

## Sources

- 1997: Donald E. Knuth:
*The Art of Computer Programming: Volume 1: Fundamental Algorithms*(3rd ed.) ... (previous) ... (next): $\S 1.2.3$: Sums and Products: $(1)$