Indexing in Math | The Ultimate Crash Course for PLEM Majors

Math Monthly · Lesson Library

Volume 5 · Indexing

Index Notation Crash Course

The Ultimate Crash Course for PLEM Majors

Study index notation with Author Jonathan David. Start with vector components and basis vectors, then learn to distinguish free indices from summed indices and use Einstein summation. Practice translating familiar vector and matrix operations into component form as preparation for tensor analysis.

Choose a lesson below.
Select its title to expand the lesson. Select it again to close it.

Index Notation and Tensor Analysis for Physics and Engineering – Mary Boas Textbook Review

Video Lesson

Watch this lesson on YouTube

Video Lesson Summary

AI summary of the video transcript.

Author Jonathan David reviews Mary L. Boas’s Mathematical Methods in the Physical Sciences as a starting point for the November 2026 Math Monthly engineering lessons. He previews index notation and Cartesian tensors and explains why students will draw on earlier coursework when they encounter this material.

  1. See how the textbook builds on earlier courses. Boas brings together material from calculus, linear algebra, differential equations, and physics. Jonathan compares opening the book to picking up where the harder chapters of those introductory textbooks leave off. He explains that the book supports several subjects and can be used chapter by chapter.
  2. Read the definition in the textbook you are using. When asked what a tensor is, start by identifying the setting and reading the author’s definition. In this lesson, the setting is Cartesian tensors, with rotations of the coordinate axes as the starting point. Jonathan also shows where he found the textbook and records the source for his notes.
  3. Use repeated indices to read a sum. The video previews Einstein summation convention, where the repeated indices in a term tell you which sums to take. In the three-dimensional Cartesian example, the indices $k$ and $l$ are summed from $1$ to $3$, while $i$ and $j$ remain free and identify the component being calculated.
  4. Start with a rank-zero Cartesian tensor. It has one component whose value stays the same when the coordinate axes rotate. In the terminology used in the lesson, this is a scalar, or invariant. Jonathan introduces this definition before discussing how tensors are used in physics and engineering.
  5. Follow the notation into later lessons. Jonathan plans to develop his own introduction to index notation and tensor operations for the Math Monthly engineering volume, starting with Boas and drawing on other textbooks as the series progresses.

“We’re looking at the theory, the definition.”

Textbook discussed
Mary L. Boas, Mathematical Methods in the Physical Sciences.

Introduction to Index Notation in 3D – Column Vector Derivation – Lesson 1

Video Lesson

Watch this lesson on YouTube

Written Lesson

Introduction to Index Notation in 3D

Column Vector Derivation

Consider a sequence or series.

Let $A=\{a_1,a_2,\ldots\}=\{a_i\}$, $i\in\mathbb{N}\equiv\mathbb{Z}^{+}$ or $A=\{a_0,a_1,\ldots\}=\{a_i\}$, $i\equiv\mathbb{Z}^{+}\cup\{0\}$.

Our indexing is $a_i$.

For a series, summing the sequence, we see

$$a_1+a_2+\cdots=\sum_{i=1}^{j}a_i=a_1+a_2+\cdots+a_j.$$

Again, our indexing is $a_i$. This is a basic relation to our quest. Let us jump right into vector analysis.

Since we are primarily sticking with engineering in this journal, we will stay within the three-dimensional realm.

Let $\mathbf{v}=(v_1,v_2,v_3)$. We consider various notations from various subjects (textbooks).

$$ \mathbf{v} =(v_1,v_2,v_3) =\begin{pmatrix}v_1\\v_2\\v_3\end{pmatrix} =\begin{bmatrix}v_1\\v_2\\v_3\end{bmatrix} =\lvert v_1,v_2,v_3\rangle =\begin{bmatrix}v_1&v_2&v_3\end{bmatrix}^{T}. $$

Depending on whether you are in standard physics, calculus, quantum mechanics, …, linear algebra and so forth, the notation changes. It is vital that you stick with the notations from the text assigned. Else, you are not doing it correctly. Very similar to using APA vs. MLA.

For the Mary Boas text, the author uses the following notation for vector.

$$\mathbf{r}=\mathbf{i}x+\mathbf{j}y+\mathbf{k}z.$$

This is a common notation from Thomas’ Calculus as well as Stewarts, but they also mix and match. Not entirely our concern at this moment. For now, we are more concerned with indexing.
Let us index the 3D vector.

$$ \begin{aligned} \mathbf{v} &=\begin{pmatrix}v_1\\v_2\\v_3\end{pmatrix} =\begin{pmatrix}v_1\\0\\0\end{pmatrix} +\begin{pmatrix}0\\v_2\\0\end{pmatrix} +\begin{pmatrix}0\\0\\v_3\end{pmatrix}\\[6pt] &=v_1\begin{pmatrix}1\\0\\0\end{pmatrix} +v_2\begin{pmatrix}0\\1\\0\end{pmatrix} +v_3\begin{pmatrix}0\\0\\1\end{pmatrix} \end{aligned} $$

NOTE $\mathbf{i}=(1,0,0)$, $\mathbf{j}=(0,1,0)$, $\mathbf{k}=(0,0,1)$. But, what if we have more than 26 letters? We must index! We are staying in 3D for this textbook though—since, this is meant for engineering (also useful for physicists).

Let a unit vector be represented as $\mathbf{e}_i$.

$$ =v_1\mathbf{e}_1+v_2\mathbf{e}_2+v_3\mathbf{e}_3 =\sum_{i=1}^{3}v_i\mathbf{e}_i. $$

Now, there is something called the Einstein Summation Convention. This means that there is no point in writing the capital sigma. Thus, we can define a vector in 3D as

$$\mathbf{v}\equiv v_i\mathbf{e}_i,\qquad i\in[1,3].$$

Since we are stationed in three-dimensional space, it is not necessary to state the interval.

Video Lesson Summary

AI summary of the video transcript.

Author Jonathan David introduces index notation for the November 2026 Math Monthly engineering volume. The lesson starts with familiar sequence subscripts and uses the same idea to express a 3D vector as a sum of its components times the standard basis vectors.

  1. Start with sequence indices. A subscript tells you which term you are referring to. Jonathan shows indexing from both zero and one and uses $\sum_{i=1}^{j}a_i$ to illustrate a finite sum. Here $i$ runs through the terms and $j$ specifies the last term.
  2. Read the notation in context. Tuples, column vectors, and basis vector expressions can all convey the same components in this example. Jonathan compares the notation across subjects and uses Mary L. Boas’s expression $\mathbf{r}=\mathbf{i}x+\mathbf{j}y+\mathbf{k}z$ as an example. He compares following a textbook’s notation to following a style convention such as APA or MLA.
  3. Build the column vector one coordinate at a time. Split the column into three vectors, each with only one potentially nonzero entry. Factoring out $v_1$, $v_2$, and $v_3$ leaves the three standard basis vectors, which can be labeled $\mathbf{e}_1$, $\mathbf{e}_2$, and $\mathbf{e}_3$.
  4. Read the repeated index as a sum. In $v_i\mathbf{e}_i$, the repeated $i$ means to add the three component–basis products. The range $i=1,2,3$ is understood once the lesson has fixed the setting as $\mathbb{R}^3$.
  5. Carry the notation into the next lesson. Jonathan proposes writing the dot product in index notation as the next exercise and previews the cross product. These examples prepare readers for the indexed expressions they will encounter in Boas’s treatment of tensor analysis.

“Many, many different ways to write the same exact thing. It just it depends on the book that you’re working in.”

The recording opens with an invitation to revisit the earlier Gustafson matrix exponential calculation and spot a misplaced entry in the $2\times2$ adjugate. Jonathan uses that example to explain how readers can take part in checking the work during the Math Monthly research sessions. He also encourages students to consult a definition when they are unsure and to practice writing mathematical explanations as carefully as they carry out calculations.

Textbook discussed
Mary L. Boas, Mathematical Methods in the Physical Sciences.

How to Write a Row Vector in Index Notation | Math Monthly Lesson 2

Video Lesson

Watch this lesson on YouTube

Written Lesson

Row Vector Derivation in 3D

Let $\mathbf{v}^{T}=\begin{bmatrix}v_1&v_2&v_3\end{bmatrix}$. Then,

$$ \begin{aligned} \mathbf{v}^{T} &=\begin{bmatrix}v_1&v_2&v_3\end{bmatrix}\\[4pt] &=\begin{bmatrix}v_1&0&0\end{bmatrix} +\begin{bmatrix}0&v_2&0\end{bmatrix} +\begin{bmatrix}0&0&v_3\end{bmatrix}\\[6pt] &=v_1\begin{bmatrix}1&0&0\end{bmatrix} +v_2\begin{bmatrix}0&1&0\end{bmatrix} +v_3\begin{bmatrix}0&0&1\end{bmatrix} \end{aligned} $$

The question here: How do we label the unit vectors? Since $\mathbf{e}_i$ is a column vector, it would be safe to assign $\mathbf{e}^{i}$ to row vectors. Thus, we continue:

November 2026 | Volume 1

$$ =v_1\mathbf{e}^{1}+v_2\mathbf{e}^{2}+v_3\mathbf{e}^{3} =\sum_{i=1}^{3}v_i\mathbf{e}^{i} $$
$$\Rightarrow\quad\mathbf{v}^{T}\equiv v_i\mathbf{e}^{i}.$$

Video Lesson Summary

AI summary of the video transcript.

Author Jonathan David builds on the column vector derivation from Lesson 1. He transposes a three-dimensional column vector, expands the resulting row vector in the standard row basis, and writes that expansion using Einstein summation convention.

  1. Transpose the column vector. The components $v_1$, $v_2$, and $v_3$ stay in the same order and are arranged horizontally in $\mathbf{v}^{T}$.
  2. Separate the components. Split the row into three rows, each carrying one component. Factoring out $v_1$, $v_2$, and $v_3$ leaves the three standard row unit vectors.
  3. Define the basis labels for this series. Jonathan uses $\mathbf{e}_i$ for the column basis vectors and $\mathbf{e}^{i}$ for the corresponding row basis vectors. The superscript in $\mathbf{e}^{i}$ is an index. He explains that other textbooks may use different labels, including $\mathbf{i},\mathbf{j},\mathbf{k}$ or their hatted forms.
  4. Read the repeated index as a sum. In $v_i\mathbf{e}^{i}$, the repeated $i$ tells you to add the three component–basis products. The shorthand expresses the same row vector as $\sum_{i=1}^{3}v_i\mathbf{e}^{i}$.
  5. Prepare for vector multiplication. The row and column expansions give the series a shared set of identities to use in later lessons on vector products, vector calculus, and tensor analysis.

“We’re creating a list of identities right now. A list of identities in 3D indexing.”

How to Write the Dot Product in Index Notation | Math Monthly Lesson 3

Video Lesson

Watch this lesson on YouTube

Written Lesson

Comparing Inner Product to Dot Product

Before we jump into the product of multiple vectors, let us product a vector with itself. We will product as standard-slang inner product and standard-slang dot product.

DEFINE Slang

In this text and all future texts, the term “slang” is implying generalized and accepted arithmetic without definition or proof. Meaning- it is common knowledge and universal—for the most part.

The Dot Product is the dot product. Basic arithmetic. However, the inner product has formality that is not being addressed—yet. We can assume that the inner product for standard vectors is similar to the dot product.

Let us perform the dot product of the vector $\mathbf{v}$ with itself.

$$ \begin{aligned} \mathbf{v}\cdot\mathbf{v} &=\begin{pmatrix}v_1\\v_2\\v_3\end{pmatrix} \cdot\begin{pmatrix}v_1\\v_2\\v_3\end{pmatrix} =(v_1)(v_1)+(v_2)(v_2)+(v_3)(v_3)\\[6pt] &=v_1^2+v_2^2+v_3^2. \end{aligned} $$

Very basic stuff. Now, let us index it.

$$ \mathbf{v}\cdot\mathbf{v} =\begin{pmatrix}v_1\\v_2\\v_3\end{pmatrix} \cdot\begin{pmatrix}v_1\\v_2\\v_3\end{pmatrix} =\left[\begin{pmatrix}v_1\\v_2\\v_3\end{pmatrix}\right] \cdot\left[\begin{pmatrix}v_1\\v_2\\v_3\end{pmatrix}\right] $$

The Math Monthly Journal

$$ \begin{aligned} &=\left[\begin{pmatrix}v_1\\0\\0\end{pmatrix} +\begin{pmatrix}0\\v_2\\0\end{pmatrix} +\begin{pmatrix}0\\0\\v_3\end{pmatrix}\right] \cdot\left[\begin{pmatrix}v_1\\0\\0\end{pmatrix} +\begin{pmatrix}0\\v_2\\0\end{pmatrix} +\begin{pmatrix}0\\0\\v_3\end{pmatrix}\right]\\[8pt] &=\left[v_1\begin{pmatrix}1\\0\\0\end{pmatrix} +v_2\begin{pmatrix}0\\1\\0\end{pmatrix} +v_3\begin{pmatrix}0\\0\\1\end{pmatrix}\right] \cdot\left[v_1\begin{pmatrix}1\\0\\0\end{pmatrix} +v_2\begin{pmatrix}0\\1\\0\end{pmatrix} +v_3\begin{pmatrix}0\\0\\1\end{pmatrix}\right]\\[8pt] &=\left[v_1\begin{pmatrix}1\\0\\0\end{pmatrix} +v_2\begin{pmatrix}0\\1\\0\end{pmatrix} +v_3\begin{pmatrix}0\\0\\1\end{pmatrix}\right] \cdot\left[v_1\begin{pmatrix}1\\0\\0\end{pmatrix} +v_2\begin{pmatrix}0\\1\\0\end{pmatrix} +v_3\begin{pmatrix}0\\0\\1\end{pmatrix}\right]\\[8pt] &=[v_1\mathbf{e}_1+v_2\mathbf{e}_2+v_3\mathbf{e}_3] \cdot[v_1\mathbf{e}_1+v_2\mathbf{e}_2+v_3\mathbf{e}_3] \quad\text{FOIL}\\[6pt] &=(v_1\mathbf{e}_1)\cdot(v_1\mathbf{e}_1) +(v_2\mathbf{e}_2)\cdot(v_2\mathbf{e}_2) +(v_3\mathbf{e}_3)\cdot(v_3\mathbf{e}_3)\\[6pt] &=(v_1\mathbf{e}_1)\cdot(v_1\mathbf{e}_1) +(v_2\mathbf{e}_2)\cdot(v_2\mathbf{e}_2) +(v_3\mathbf{e}_3)\cdot(v_3\mathbf{e}_3) \quad\text{COMMUTE}\\[6pt] &=v_1v_1(\mathbf{e}_1)\cdot(\mathbf{e}_1) +v_2v_2(\mathbf{e}_2)\cdot(\mathbf{e}_2) +v_3v_3(\mathbf{e}_3)\cdot(\mathbf{e}_3)\\[6pt] &=v_1v_1\mathbf{e}_1\cdot\mathbf{e}_1 +v_2v_2\mathbf{e}_2\cdot\mathbf{e}_2 +v_3v_3\mathbf{e}_3\cdot\mathbf{e}_3 \end{aligned} $$

Now, we can match up to $v_1^2+v_2^2+v_3^2$. But, before we do this, let us create a new identity. We may need it!

$$ v_1v_1\mathbf{e}_1\cdot\mathbf{e}_1 +v_2v_2\mathbf{e}_2\cdot\mathbf{e}_2 +v_3v_3\mathbf{e}_3\cdot\mathbf{e}_3 =\sum_{i=1}^{3}v_i^2\mathbf{e}_i\cdot\mathbf{e}_i $$

This may not be good enough for what we are doing here. We should expand it to 9 elements. Recall

$$ \mathbf{e}_i\cdot\mathbf{e}_j =\begin{cases} 0,&i\ne j,\\ 1,&i=j. \end{cases} $$
$$ \begin{aligned} &v_1v_1\mathbf{e}_1\cdot\mathbf{e}_1 +v_2v_2\mathbf{e}_2\cdot\mathbf{e}_2 +v_3v_3\mathbf{e}_3\cdot\mathbf{e}_3\\[6pt] ={}&v_1v_1\mathbf{e}_1\cdot\mathbf{e}_1+0+0+0 +v_2v_2\mathbf{e}_2\cdot\mathbf{e}_2\,0+0+0 +v_3v_3\mathbf{e}_3\cdot\mathbf{e}_3 \end{aligned} $$

November 2026 | Volume 1

$$ \begin{aligned} ={}&v_1v_1\mathbf{e}_1\cdot\mathbf{e}_1 +v_2v_2\mathbf{e}_1\cdot\mathbf{e}_2 +v_3v_3\mathbf{e}_1\cdot\mathbf{e}_3\\[6pt] &+v_1v_1\mathbf{e}_2\cdot\mathbf{e}_1 +v_2v_2\mathbf{e}_2\cdot\mathbf{e}_2 +v_3v_3\mathbf{e}_2\cdot\mathbf{e}_3\\[6pt] &+v_1v_1\mathbf{e}_3\cdot\mathbf{e}_1 +v_2v_2\mathbf{e}_3\cdot\mathbf{e}_2 +v_3v_3\mathbf{e}_3\cdot\mathbf{e}_3. \end{aligned} $$
$$\sum_{i=1}^{3}\sum_{j=1}^{3}v_i^2\mathbf{e}_i\cdot\mathbf{e}_j$$

Before we add something like this to an identity—meaning, we are going to drop the summation symbols for Einstein Summation Convention, we have established ground rules. There are conditions for the Einstein Summation Convention to be utilized. This is what our journey in indexing will lead to. Then, we will venture into Tensor Analysis.

For Example,

DEFINE Vector Product with Itself

$$ \mathbf{v}\cdot\mathbf{v} \equiv v_i^2\mathbf{e}_i\cdot\mathbf{e}_i \quad\text{Single Summation} $$ $$ \mathbf{v}\cdot\mathbf{v} \equiv v_i^2\mathbf{e}_i\cdot\mathbf{e}_j \quad\text{Double Summation} $$

We can define that. It may or may not be useful in our journey of proofs/showing identity relations.

Video Lesson Summary

AI summary of the video transcript.

Author Jonathan David takes the dot product of a three-dimensional vector with itself and uses the calculation to show how a double sum reduces to a single sum.

  1. Start with the component calculation. The self-dot product is $v_1^2+v_2^2+v_3^2$, the squared length of the vector.
  2. Write each vector in the standard basis. Distribute the dot product over the two basis expansions and move the scalar coefficients outside each basis-vector dot product.
  3. Account for all nine terms. The six pairings of different basis vectors are zero. Each of the three matching pairs has a basis-vector dot product of one, so only the squared components remain.
  4. Extend the calculation to two vectors. Jonathan previews how multiplying corresponding components of $\mathbf{u}$ and $\mathbf{v}$ replaces squaring the components of one vector.
  5. Set the rules before dropping the summation signs. The lesson keeps the sums explicit while developing the rules for index notation. It previews the next discussion of inner products and later work on vector identities such as the scalar triple product.

“There are conditions for the Einstein Summation Convention to be utilized.”