Writing a linear program in standard form Usually I have been asked to write problems in standard form that have inequalities involved. However, this problem has none and I was wondering if anyone had insight on how to go about solving it.
Consider this system:
$$
e_{i} = b_{i} - \sum_{j=1}^na_{i,j}x_{j} \\(i=1,2,...m) 
$$
where $a_{i,j}$ $(1 \leq i \leq m\,,\ 1 \leq\  j \leq n)$ and $b_{i}$ $(1 \leq i \leq m)$ are given. The problem is to find an assignment of values to the variables $x_{1}, ...,x_{n}$ that minimizes max$|e_{j}|$.
Express this problem as a linear program in the standard form.
So, there are no inequalities and both $i$ and $j$ are both always positive so I don't know where to start introducing new variables.
 A: Let $e=\begin{bmatrix}e_1\\\vdots\\ e_m\end{bmatrix}$, $b=\begin{bmatrix}b_1\\\vdots\\ b_m\end{bmatrix}$, $A=\begin{bmatrix}a_{11}&\ldots&a_{1n}\\\vdots\\ a_{m1}&\ldots&a_{mn}\\\end{bmatrix}$, $x=\begin{bmatrix}x_1\\\vdots\\ x_n\end{bmatrix}$
Then you can equivalently right your system as:
$$e=b-Ax$$ 
Let, scalar $t$ be the value of $\max_j |e_j|$, then you will have:
$$e_j\leq t~~\forall j=1,\ldots,m\\ -e_j\leq t ~~\forall j=1,\ldots,m$$
Now let $\bf{1}= \begin{bmatrix}1\\\vdots\\ 1\end{bmatrix}$ be a vector of $m$ stackes $1$'s, then these constraints can be written as:
$$e\leq t1\\ -e\leq t1$$
where $\leq$ is the element-wise less than/or equal. Equivalently you have: 
$$b-Ax\leq t1\\ Ax-b\leq t1$$
the linear program that you are looking for will be:
$$\max_{x,t} ~~t\quad\text{subject to}\\b-Ax\leq t1 \\ Ax-b\leq t1$$
$$\max_{x,t} ~~t\quad\text{subject to}\\-Ax-t1\leq -b \\ Ax-t1\leq b$$
If you want to standardize it further, do the following: let $z=[x^T~~t]^T$, $c=[\underbrace{0\ldots 0}_{n \text{ times}}~~1]^T$, $d=[-b^T~~b^T]^T$, $H=\begin{bmatrix}-A& -1\\A& -1\\\end{bmatrix}$, now it can be written as:
$$\max_{z} ~~c^Tz\quad\text{subject to}\\Hz\leq d$$
