Bookmark and Share

Testing math content rendering with MathJax. Test cases are originally copy from Gautam Iyer's test cases, and then gradually modified and suplemented.

Inline math

This is an equation: $E = m c2$.


Theorem (Mean Value Property). Let $\Omega \subset \mathbb{R}^3$ be a domain, and $u$ is harmonic in $\Omega$ (i.e. $u = 0$ in $\Omega$). Suppose $B$ is a ball of radius $R$ and center $x_0$ that is completely contained in $\Omega$. Then $$ u(x_0) = \frac{1}{4 \pi R2} \int_{\partial B} u \, dS $$



Subscripts: ${{a} {1}}^{{b} {1}}$

Curly braces

The curly brace is a special character in LaTeX. Therefore, to display it, it needs to be escaped by a backslash. The backslash that is not part of the original LaTeX keyword such as \right needs to be escaped. So, at the end you need \\} to display a closing curly brace:

\left \\{X_1, X_2, \cdots, X_N \right \\} displays as $\left \{X_1, X_2, \cdots, X_N \right \}$.

Multi-line equations

$$ \begin{equation} \label{eq1} \begin{split} A & = \frac{\pi r2}{2} \\ & = \frac{1}{2} \pi r2 \end{split} \end{equation} $$

$$ \begin{align} 2x - 5y &= 8 \\ 3x + 9y &= -12 \end{align} $$

$$ \begin{align} x&=y & w &=z & a&=b+c\\ 2x&=-y & 3w&=\frac{1}{2}z & a&=b\\ -4 + 5x&=2+y & w+2&=-1+w & ab&=cb \end{align} $$

Rotational Symmetry of the Laplacian

Let $T:\Real2 \to \Real2$ be a rotation, and $u$ be any function. Then $$ \Laplace (u \circ T) = (\Laplace u) \circ T $$ This is also true in higher dimensions.

Given that the Laplacian has this rotational symmetry, one might expect a nice cancellation / compact formula for the Laplacian in polar coordinates. We try this next.

The Laplacian in Polar Coordinates

Let $\hat x$, $\hat y$ be the unit vectors in the $x$ and $y$ direction respectively, and let $\hat r$ and $\hat \theta$ be the unit vectors in the $r$ and $\theta$ direction respectively. Explicitly, $$ \hat x = \begin{pmatrix}1\cr 0\end{pmatrix},\quad \hat y = \begin{pmatrix}0\cr1\end{pmatrix},\quad \hat r = \frac{1}{r}\begin{pmatrix}x\cr y\end{pmatrix},\quad \hat \theta = \frac{1}{r}\begin{pmatrix}-y\cr x\end{pmatrix}. $$

First we compute that $$ \Grad u = \partial_x u \hat x + \partial_y u \hat y = \partial_r u \hat r + \frac{1}{r} \partial\theta u \hat \theta. $$ Now we compute $$ \Laplace u = \Div \Grad u = (\Grad \partial_r u) \cdot \hat r + \partial_r u (\Div \hat r) + (\Grad \frac{1}{r} \partial\theta u) \cdot \hat \theta + 0 = \partial_r2 u + \partial_r u (\Div \hat r) + \frac{1}{r2} \partial_\theta2 u. $$

To finish the calculation, we only need to compute $\Div \hat r$. We do this as follows $$ \Div \hat r = \Div \left( \frac{1}{r} \begin{pmatrix} x \cr y \end{pmatrix} \right) = \frac{2}{r} + \Grad\left( \frac{1}{r} \right) \cdot \begin{pmatrix} x \cr y \end{pmatrix} = \frac{2}{r} - \frac{1}{r} = \frac{1}{r}. $$ Substituting back gives $$ \Laplace u = \partial_r2 u + \frac{1}{r} \partial_r u + \frac{1}{r2} \partial_\theta2 u. $$

Basic tests to check interference with markdown.

Sub and super scripts.

Here's a superscript without math: ab, a_b. Here they are with math:

  1. One backslash: ( ab, a_b ).

  2. Two backslashes: \( ab, a_b \).

  3. Three backslashes: \( ab, a_b \).


{x} produces {x} in normal mode. In math mode it will mess things up! However, smileys need whitespace before and after to be recognized. So while $ {x} $ will mess things up, ${x}$ will be OK and produce ${x}$.

Checking if * and _ mess up math

The code

a * b + b * c \qquad
a_b + b_c, \qquad
a^{b + c}

produces $$ a * b + b * c \qquad a_b + b_c, \qquad a^{b + c} $$

MathJAX equation reference test

Here is a labeled equation: \begin{equation}x+1\over\sqrt{1-x2}\label{ref1}\end{equation} with a reference to ref1: \ref{ref1}, and another numbered one with no label: $$x+1\over\sqrt{1-x2}$$ This one uses \nonumber: \begin{equation}x+1\over\sqrt{1-x2}\nonumber\end{equation}

Here's one using the equation environment: \begin{equation} x+1\over\sqrt{1-x2} \end{equation} and one with equation environment: \begin{equation} x+1\over\sqrt{1-x2} \end{equation*}

This is a forward reference [\ref{ref2}] and another \eqref{ref2} for the following equation: \begin{equation}x+1\over\sqrt{1-x2}\label{ref2}\end{equation} More math: $$x+1\over\sqrt{1-x2}$$ Here is a ref inside math: (\ref{ref2}+1) and text after it.

\begin{align} x& = y_1-y_2+y_3-y_5+y_8-\dots && \text{by \eqref{ref1}}\cr & = y'\circ y^* && \text{(by \eqref{ref3})}\cr & = y(0) y' && \text {by Axiom 1.} \end{align}

Here's a bad ref [\ref{ref4}] to a nonexistent label.

An alignment: \begin{align} a&=b\label{ref3}\cr &=c+d \end{align} and a starred one: \begin{align} a&=b\cr &=c+d \end{align}

Custom macro tests

This tests some of my predefined macros.


$$ a < b, \quad a \leq b \quad a \geq b \quad a > b. $$


Inline limits: $\displaystyle \lim {x \to a} \frac{1}{x}$, $\displaystyle \max {0, 1}$.

Displayed \displaystyle {\operatorname{ess} \sup} _{x \in \Real} produces $$ \displaystyle {\operatorname{ess} \sup} _{x \in \Real} $$


Subscript _{xxx} needs to be escaped \_{xxx}

Since pairs of _ denotes emphasis in Markdown syntax, it needs to be escaped:

&= \frac{1}{N} \sum\_{i=1}^N \left( X\_i - \bar{X}\_N \right)^2 \\\\
&= \frac{1}{N} \sum\_{i=1}^N \left( X\_i^2 - \bar{X}\_N^2 \right)~,

$$ \begin{align} \hat\sigma2_\mathrm{N} &= \frac{1}{N} \sum_{i=1}^N \left( X_i - \bar{X}_N \right)2 \\ &= \frac{1}{N} \sum_{i=1}^N \left( X_i2 - \bar{X}_N2 \right)~, \end{align} $$

Or, use a white space in front of _:

&= \frac{1}{N} \sum\_{i=1}^N \left( X\_i - \bar{X}\_N \right)^2 \\\\
&= \frac{1}{N} \sum\_{i=1}^N \left( X\_i^2 - \bar{X}\_N^2 \right)~,

$$ \begin{align} \hat\sigma2 \mathrm{N} &= \frac{1}{N} \sum {i=1}^N \left( X i - \bar{X} N \right)2 \\ &= \frac{1}{N} \sum {i=1}^N \left( X i2 - \bar{X} _N2 \right)~, \end{align} $$

blog comments powered by Disqus