March 2006 Archives

As you can probably tell, my blog has been rather silent recently. I’ve been teaching and working on graduating. My defense date is finally set: June 14^th. I’ve written two chapters already. (The first one has been published in Bioinformatics.) I just have one more chapter to write. I think that I’m going to bite the bullet and write my dissertation in . I’ll probably be done with my dissertation in April or early May, and just have to wait for my committee to get back in to the country before I defend.

I am a Graduate Lab Assistant this semester. I teach two sections of BIOL 1108L Thursday afternoons (intro biology lab for science majors, second semester). The lab used to consist mostly of students studying taxonomy and classifying specimens. However, we have a new lab manual this year. The new lab is inquiry based, and we are trying to teach the students to think like scientists. There are three weeks of introduction then two five-week sections in which groups of four students perform two experiments of their own design. The first section is about lake ecosystems and the second section is about forest ecosystems. The lab mostly focuses on ecology and organismal biology for practical reasons. I would love to see a stronger emphasis on evolutionary biology, specifically evolutionary mechanisms, but with only two hours per lab the students don’t have the time to do a manipulative evolutionary experiment.

The five-week sections look like this:

They have to turn in their assignments 72 hours before the next class, and I have to grade and return their assignments before the next class.

This week is week 3 of the forest section. I’m cutting a couple of exercises and replacing them with an exercise on mathematical modeling. There is nothing at all in the lab manual touching on biology’s use of mathematics. I feel strongly that biologists are failing our undergrads for the most part ignoring mathematical biology in their educations. We require them to take calculus, but we don’t integrate it into biology classes to the same level that it is integrated into biological sciences.

In this challenge, I’m going to expose them to a model of exponential population growth, so they will see how a population density model works. The challenge for them is to come up with a model of population growth for predator-prey interaction. Each group is going to be provided a desktop so they can implement and explore their models in Excel. I’m hoping that it will go well.

Well, we’ll see come Thursday.

I’ve been using TeXnicCenter over the past month to compose my LaTeX files. I really like it the only problem is it comes with limited support for interacting with Acrobat. Acrobat doesn’t play nice with other software programs because it locks any pdf that it is viewing, which prevents any other program from accessing the same file. Thus when I try to modify the pdf using TeXnicCenter, pdflatex fails.

The current solution is to issue DDE commands to get Acrobat to close the pdf before it is written to and reopen it after it is finished. However, because the Acrobat DDE interface it is difficult to make Acrobat open the pdf to the place you want it to. For example, say you want to modify line 400 in your LaTeX file and look at the results in the pdf. Using the current solution, every time you open the pdf, you’d have to scroll down several pages until you find the one where line 400 is.

The ideal solution would be for TeXnicCenter to direct the pdf to open to the exact spot where line 400 is. However, the association of tex lines with pdf pages is not an easy one to estimate. So, I came up with an alternative solution that should work for most cases.

My solution, pdfpos, is to record on close where the pdf is currently being viewed and then scroll to that location when the new version is opened. Therefore, if you are currently viewing the location you wish to modify the new pdf will scroll to that location when it is opened.

My program uses the OLE interface to Acrobat which is more powerful than the DDE interface. I wrote pdfpos in Perl, and thus it requires Perl to run. (Get ActivePerl.) In addition you need the modules Win32::OLE and Win32::GuiTest. Visual Basic would offer a more natural implementation of this technique, but I don’t know VB.

To integrate pdfpos into TeXnicCenter simply create a new build profile, copied from one of your existing pdf profiles. Then change the viewer from “path\to\acrobat.exe” to “path\to\pdfpos.bat”. Then under view and forward search select command line argument and enter view "%bm.pdf". Similarly under close select command line argument and enter close "%bm.pdf". (Download an importable TeXnicCenter build profile.)

Download pdfpos.bat.

The code is below the fold.