My knack for checking for software releases on the release day shines again today. Several months ago I had heard a rumor about the Enthought Python Distribution (EPD) being developed for the Mac. Between writing my dissertation and packing up my family for our upcoming move, I hadn’t thought about it again… until today. And of course, today this announcement was posted. So I’m currently pulling down a 200 MB installer package and I’ll post again soon once I have a chance to tinker.
]]> http://dawes.wordpress.com/2008/07/02/epd-for-os-x-public-beta-available/feed/ adawes http://dawes.wordpress.com/2008/07/01/subversion-and-mac-bundles/ http://dawes.wordpress.com/2008/07/01/subversion-and-mac-bundles/#comments Tue, 01 Jul 2008 19:27:37 +0000 adawes http://dawes.wordpress.com/?p=151

I wrote earlier about using SVN and LaTeX for document management. While this has continued to work well for me, I did discover an issue with some document files generated by mac apps. It has become common to save documents as “bundles,” which from the filesystem perspective are directories but look like documents in Finder. Unfortunately, these bundles delete or overwrite any .svn folders they contain, so if you put an iWork document, or in my case a DataGraph plot, under revision control in SVN, you will get an error after you save it the next time:

svn: Working copy ____ is missing or not locked.”

I found a handy Ruby script that has a built-in list of extensions to look for (iWork files among others). I simply add dgraph to the list of extensions and it automatically downloads and replaces the missing .svn folders.

This is somewhat off topic, but I felt it was a good example of a technology mashup that serves those of us who would like to drive less. Of course it is most helpful to those of us who have a chance to choose a place to live based on how walkable that place is.

Walk Score is a hybrid google business map that tabulates the general distance to a wide range of typical business types.

The walk score then rates how reasonable it is to live car-lite, or survive and walk everywhere. My first three data points actually reflect their rankings very well. My childhood home gets a 57 or “some walkable locations.” I found it easy to bike around in the years before I had a driver’s license, but I wouldn’t want to live there without a car as an adult. My current house gets a 38: “not walkable.” I find this generally true from the perspective of errands. However, our proximity to the park easily offsets the requirement of driving to the grocery store. A lot of places are an easy bike-ride: the farmer’s market, downtown, brightleaf square, American Tobacco Campus, Duke East campus, and 9th street (to name a few). Fortunately Durham is gaining ground in the bike-friendly world.

Finally, our new house, in Forest Grove, gets a 74: “Very walkable.” Woohoo! Of course we knew this about the house; it was part of what we liked best. And it certainly validates the algorithm that walkscore uses.

It’s been a little while since CLEO and my last post. Things have been hectic to say the least. The past few weeks included:

• a trip to Forest Grove to look for houses
• a week at home with Carter (17 mo. old) while Leslie went to Volleyball tournament
• showing our house twice
• receiving two offers on our house
• signing one of them

We’ve had the home inspection and are waiting on the “Requested Repairs” paperwork, otherwise things are looking good.

“What does this have to do with Waves and Optics?” you ask. Well that is the course that I am responsible for this fall. I’m also teaching Geometrical Optics I and Workshop Physics (calc-based intro) but those courses are already put together.

Waves and Optics with Prof. Dawes (aka PHY 332)… what will it be, what wonders of physics will be explored and experienced by students in my class? I am leaning towards using H. J. Pain’s “The Physics of Vibrations and Waves” which starts with simple harmonic motion, builds on the wave equation from SHM, and then branches out to sound, water, and EM waves. The latter parts of the book are primarily optics: diffraction, interference, scattering etc. I really want to work in some current research results, especially since things like slow- and fast-light naturally come out of a discussion of wave physics. Another neat topic in the news lately are negative-index metamaterials. Of course there are also my own research interests: pattern formation, and nonlinear optics.

I don’t think I’ll have trouble coming up with current material, but I’m happy to hear other suggestions on recent topics that would fit into a course on waves and optics. I’m sure there are plenty that I’m not thinking of at the moment. Especially welcome are suggestions from any of you experts outside my subfield of optics.

I’m currently attending the CLEO/QELS conference in San Jose, CA. Yesterday there was a symposium honoring the late Ted Maiman, who was the first person to successfully build and operate a Laser. After the symposium, I was doing some follow-up reading on Wikipedia and discovered that the timing of this memorial session was very coincidental. Ted passed away a year ago today, and his famous invention, according to his notebook, took place almost exactly 48 years ago (May 16, 1960). For the Conference on Lasers and Electro-Optics to take place between these dates is highly appropriate, and it has been interesting to consider the conference within the context of the historical origins of the laser. Hopefully Ted would be impressed with what we have all come up with using his invention.

For anyone curious about the process of managing a LaTeX document with the Subversion (SVN) version control system, I have to highly recommend it. Now that my dissertation is officially finished, I have a bit of time to explain the process I used to back-up, archive, and otherwise manage the beast.

The three major components of the computing end of my dissertation were Textmate (and the LaTeX and SVN bundles), Google Code, and of course LaTeX.

I set up a project at Google Code (free hosting for < 100 MB) which includes SVN access, a project wiki, and a nice web interface. Almost all of my LaTeX was written in the Textmate editor simply because it has a lot of useful macros and plugins including a feature that lets you drag and drop images to make figures (it automatically inserts the \begin{…} \end{…} code and has lots of other type-savers.

Another advantage of Textmate is built-in SVN support. The above screen capture shows the result of an SVN “commit” which happened to be my 200th revision. In the end, I had 202 revisions, but the last two were trivial (adding an updated figure file, and a little typo).

There is a fair bit of documentation out there about how to do each of these things individually, so I won’t reiterate any of the instructions here. I mainly want to suggest this approach to anyone who is managing a large document, or especially sharing the responsibility with multiple authors.

I don’t know if this was an intended use for Google Code, but it certainly serves the computing community well. If you are interested, the repository for my thesis (i.e. the source code) is available at http://code.google.com/p/dawes-phd-thesis/

It’s finally done. I submitted the final version of my dissertation[PDF] to the Graduate School. Naturally, that means I’ll be headed to the beach tomorrow for a weekend of camping in the sand.

Not that I expect anyone to, but you’re welcome to read it, just as long as you don’t tell me about any typos… at least not until next week sometime

It’s official. I successfully defended my dissertation today. The process was surprisingly enjoyable. After very little sleep, I still managed to give a good presentation and field a bunch of good questions. My committee was genuinely interested in my work and gave me lots of new ideas on where I can take the research.

I have some corrections to make in my thesis, but otherwise I can relax a little bit. I’m glad I’ll have a chance to write up some of my newest results and maybe get another paper out. That should make the transition to a new career at Pacific a little more productive. I’ll post once my thesis is finalized and available online.

-Dr. Dawes

I’ve made my last revisions prior to handing my dissertation over to my committee members tomorrow (ok, technically that’s just later today). My defense is scheduled for 9am on April 11th so I’ve got just over a week to put together my presentation. After writing 160 pages on my research, I think 20 slides will be straightforward. And no, I won’t just put 8 pages of information on each slide, that would be science-talk suicide.

As you can imagine, during the depths of my thesis-writing experience, I don’t have a whole lot of time for anything except… well, writing my thesis. Sometimes, especially in the sciences, as you are writing the paper, the data is still coming in and ideally still getting better. Lucky for me that was the case today. The new data meant new analysis, which for my project means more python code.

It’s never happened to me before, but today was a notable day in my evolution as a programmer: I wrote my first non-trivial program, from an empty file, to working code, in one iteration. Now, of course, I can write a “hello world” program in fortran or c++ without debugging (I might have to change linker flags, but that doesn’t count), but today it was a real program, doing real things… things that were non-trivial in a programming sense.

If you want to know the details, the program runs through an array (presumably time-series data, but it could be anything), checks for level-crossings, interpolates between sampled data points, and returns the interpolated level-crossing as a list of x,y values to make plotting easy. I had level-crossing code before, so it wasn’t entirely from scratch, but it was enough to be a real “oh, wow, it worked right the first time!” moment. I know for sure it would have taken me the whole day to work this up in c++. And that is why I love python.