Creating a handout from beamer slides

I’m about to head off on a speak­ing tour to Europe (more on that in another post) and one of my hosts has asked for my pow­er­point slides so they can print them. They have made two false assump­tions: (1) that I use pow­er­point; (2) that my slides are sta­tic so they can be printed.

Instead, I pro­duced a cut-​​down ver­sion of my beamer slides, leav­ing out some of the ani­ma­tions and other fea­tures that will not print eas­ily. Then I pro­duced a pdf file with sev­eral slides per page. Con­tinue reading →

Removing white space around R figures

When I want to insert fig­ures gen­er­ated in R into a LaTeX doc­u­ment, it looks bet­ter if I first remove the white space around the fig­ure. Unfor­tu­nately, R does not make this easy as the graphs are gen­er­ated to look good on a screen, not in a document.

There are two things that can be done to fix this prob­lem. Con­tinue reading →

Bare bones beamer

Beamer is far and away the most pop­u­lar soft­ware for pre­sen­ta­tions amongst researchers in math­e­mat­ics and sta­tis­tics. Most con­fer­ence and sem­i­nar talks I attend these days use beamer. Unfor­tu­nately, they all look much the same. I think peo­ple find beamer themes too hard to mod­ify eas­ily, so a small num­ber of tem­plates get shared around. Even the oth­er­wise won­der­ful LaTeX Tem­plates site has no beamer examples.

The beamer user guide explains how to make changes but it is not for the faint-​​hearted (although it is a fan­tas­tic resource once you have some expertise).

So I thought it might be use­ful to pro­duce a very sim­ple beamer tem­plate that is easy to extend and mod­ify. Con­tinue reading →

Making a poster in beamer

This week, I made my first poster. Although I’ve been an aca­d­e­mic for more than 20 years, I’ve never had to make a poster before. Some of my coau­thors have made posters about our joint research, and two of them have even won prizes (although I can’t take any credit for them). But this week, our depart­ment is dis­play­ing posters from all research staff about our recent work.

Here is my poster (click for pdf ver­sion):

It was done using beamer which turns out to be as good for posters as it is for slides. I used the beam­er­poster pack­age which comes with a few themes. None of the themes were quite what I wanted, so I adapted one. Here it is in case any­one else wants to use it. I’ve also made a tem­plate based on the poster above.

LaTeX templates for Monash

I have updated my LaTeX tem­plates for use at Monash so they no longer depend on HyTeX. Tem­plates are pro­vided for pro­duc­ing a work­ing paper and a PhD the­sis in the Depart­ment of Econo­met­rics & Busi­ness Sta­tis­tics at Monash Uni­ver­sity. Peo­ple at other uni­ver­si­ties are wel­come to adapt the tem­plates for their own institutions.

Giving a research seminar

An expanded ver­sion of this post is avail­able in my arti­cle on “Giv­ing an aca­d­e­mic talk”.

With con­fer­ence sea­son almost upon us, it is timely to dis­cuss what makes a good con­fer­ence pre­sen­ta­tion. Here is a sug­gested structure.

1. A moti­vat­ing exam­ple demon­strat­ing the prob­lem you are try­ing to solve.
2. Explain exist­ing approaches to the prob­lem and their weaknesses.
4. Show how your ideas solve the problem/​example you started with.

That won’t nec­es­sar­ily work for every talk, but it is a good place to start. In par­tic­u­lar, begin­ning with a moti­vat­ing exam­ple is much bet­ter than set­ting up the prob­lem algebraically.

Fur­ther suggestions:

• Use beamer with this tem­plate.
• Use a max­i­mum of 20 slides for a 20 minute con­fer­ence presentation.
• Assume the audi­ence knows about what you did at the start of your research in this area. That is, prob­a­bly not much. You can assume stan­dard mate­r­ial taught to under­grad­u­ates (regres­sion, ARIMA mod­els, etc.), but don’t assume they already know what you have spent long hours learn­ing on your own.
• Give only the most nec­es­sary math­e­mat­i­cal details. Peo­ple do not quickly absorb math­e­mat­i­cal equa­tions so don’t give any more than you have to. Never give proofs.
• When you do include some alge­bra, define all terms used. Why make the audi­ence guess?
• Use graphs instead of tables where possible.
• Where pos­si­ble, let the graph fill the slide. The com­mand \fullwidth{file} is useful.
• Use \begin{block}...\end{block} to high­light equa­tions or impor­tant information.
• Use \structure{...} to high­light head­ings on slides.
• Do not use equa­tion num­ber­ing, fig­ure num­ber­ing, etc. The lis­ten­ers can’t go back and see which one you are refer­ring to.
• At the bot­tom of the last slide, give your web­site or email address for peo­ple to con­tact you if they want to read the paper or down­load your R code.
• Go through all your slides and see what you can remove. Less text is better.
• Go through all your slides again and check that the titles are empha­siz­ing the right thing. Fix them where necessary.
• Go through all your slides again just to make sure you can’t see any­thing that could be improved.