# Upright greek letters

I was wondering if anyone knew how I could produce upright greek letters on stats.SE? In my documents, I usually do this using the $\LaTeX$ upgreek package.

### Edit

I am uploading an image of what I hope the upright greek letters look like. The $\LaTeX$ code to produce this document is included below.

\documentclass{article}
\usepackage[table]{xcolor}
\usepackage{amsmath, upgreek, booktabs, tabularx, colortbl}
\usepackage[active, tightpage]{preview}
\begin{document}
\begin{table}[htbp]
\begin{preview}
\begin{tabular}{ccc}
\toprule
& Deterministic & Non-deterministic \\
\midrule
Scalar & $\alpha, \beta, \gamma$ & $X, Y, Z$ \\
Vector & $\boldsymbol{\alpha}, \boldsymbol{\beta}, \boldsymbol{\gamma}$ & $\boldsymbol{X}, \boldsymbol{Y} \boldsymbol{Z}$ \\
Matrix & \cellcolor{blue!25}$\boldsymbol{\upalpha}, \boldsymbol{\upbeta}, \boldsymbol{\upgamma}$ & $\mathbf{X}, \mathbf{Y}, \mathbf{Z}$ \\
\bottomrule
\end{tabular}
\end{preview}
\end{table}
\end{document}

-
You might check the MathJax website and forum. Out of purely idle curiosity: Is the motivation behind your question some aesthetic preference or is there a (statistical) communication issue at play here that you're seeking to resolve? – cardinal Nov 30 '12 at 13:21
@cardinal Thanks. I will look up the MathJax website. The motivation is statistical communication. I follow slightly atypical notation (but found in some multivariate statistics textbooks) to denote parameter matrices by upright bold, parameter vectors by regular (italic) bold. I distinguish deterministic and non-deterministic quantities using cases -- uppercase (greek or roman) for non-deterministic and lower case for deterministic. – tchakravarty Nov 30 '12 at 14:04
@cardinal, some scientific standards organizations (e.g. NIST, IUPAC) stipulate italic font for letters representing variables or functions, & roman font for letters representing mathematical constants, descriptive terms, or SI prefixes. So an upright pi for the ratio of the circumference of a circle to its diameter, but an italic pi for the parameter of a binomial distribution. – Scortchi Dec 5 '12 at 15:07
@Scortchi: (+1) This is a very interesting observation. It seems to be of a different motivation than fg nu's, though. At any rate, these are manifestly not conventions in (the vast majority of) mathematical writing (nor, that I am aware, in its brethren in statistics, engineering, etc.) – cardinal Dec 5 '12 at 15:23
@scortchi Thanks for the comments. As cardinal has noted, there are, regrettably, few common notational conventions in statistics, although in econometrics, there has been some attempt at unification. The best thing to do appears to be to develop a reasonable personal system of notation, and stick to it. – tchakravarty Dec 5 '12 at 15:39
@cardinal, this convention seems to be rarely observed for mathematical constants represented by Greek letters, occasionally observed for mathematical constants represented by Latin letters (e.g. e for the base of natural logarithms), & often observed for operators (e.g. d for the differential operator). Geography as well as discipline may play a part - & I'd bet what's easiest in $\LaTeX$ too. – Scortchi Dec 5 '12 at 17:35

## 1 Answer

I think you have to use unicode within the $\LaTeX$ code:

${\bf \unicode[Times]{x3B1}}$

${\bf \unicode[Times]{x3B2}}$

${\bf \unicode[Times]{x3B3}}$

will result in

${\bf \unicode[Times]{x3B1}}$

${\bf \unicode[Times]{x3B2}}$

${\bf \unicode[Times]{x3B3}}$

See the Wikipedia page for the Greek Alphabet for the unicode for each letter.

Note that you can use regular $\LaTeX$ commands for upright uppercase Greek letters with \rm. For example, using:

${\rm A, B, \Gamma}$

results in

${\rm A, B, \Gamma}$

-
I have edited my question to show what I I hope the end result looks like. I must admit, I am not a huge fan of mixing unicode with LaTeX. – tchakravarty Nov 30 '12 at 14:44
No, me neither! But I think with MathJax it's the only way to get the non-italic lower case Greek letters. I'd love to be proven wrong though! – smillig Nov 30 '12 at 15:11