# What is the best way to prepare a two-column paper in LaTeX?

Scientific papers are usually set in a two-column lay-out. Over years I have tried to find a good solution for typing such documents until I arrived at this website.
All my efforts until now did not deliver acceptable results and included a lot of hacking. The most troublesome thing is setting figures. Many graphs do not need to span over two columns but would nicely fit into one column. Yet, I could not find an acceptable way to do that in LaTeX.

Therefore the question:

What is the best way to prepare a two-column paper in LaTeX?

-
I'm confused. Why does \begin{figure}...\end{figure} not work for you for setting figures in only a single column? – TH. Oct 9 '10 at 16:32

The best solution is to use the class file provided by the publisher of the journal to which you intend to submit the paper. Ie, use revtex for APS and AIP journals, elsarticle for Elsevier and so on.
You need not worry about making the paper look perfect: this is the job of the copyeditors employed by the journal (or working for a copyediting company in, eg, India). In fact, by doing any latex hacking you are generally making things more difficult for the publisher because such things can interfere with the publisher's automated workflow and require additional manual intervention (and therefore delay publication of your article). Many publishers thus refuse to accept manuscripts that use any kind fancy tex. For example, APS forbids use of \def, \gdef, \edef as well as \if, \footnotemark, creation or modification of tex counters, \newfont, \symbol, generalized lists (\begin{list}), amsmath commands \genfrac, \smash and many others. They allow certain other constructs but warn that they will likely be ignored in production, such as \labelstyle and \tabbing.
Wow. I'd be mad as hell if people edited how my papers look. I guess I'm fortunate that my field doesn't typically bother with journals and conference proceedings publications are of the pdf, not the tex. (I'm not sure if it was an omission in your list, but I'd use \xdef out of spite if I had those requirements.) – TH. Oct 9 '10 at 16:35