Search Results: "Dirk Eddelbuettel"

2 January 2007

Dirk Eddelbuettel: And another New Year's Day 5k run in Lincoln Park

For the fifth time in a row, I started yet another year with 5k race by participating in the 22th New Year's Day 5k at the lakefront. The race had decent conditions. Chicago is currently having unusually mild weather, which is an extra bonus. It was well over 50 degrees yesterday, but a little cooler around fourty this morning, and unfortunately overcast. And luckily fairly little wind as it can be unplesant at the lake. Following some mild training runs both Saturday and Sunday, I took it relatively easy and finished around 20:15 or 20:20 -- I forgot to stop the Garmin at the finish, and the official results are not yet published. Not a bad time given that I haven't really trained since the last marathon.

30 December 2006

Dirk Eddelbuettel: Version 0.2.3 of digest

A new version of digest has been sent to CRAN and should appear on the CRAN mirrors in a few days; in the meantime you can fetch it from here too. This version is a simple maintenance release that adds a call of path.expand() around the optional file argument so that things like ~ get expanded by R. My thanks to Jarek Tuszynski and Tom Bloemberg for the suggestion.

8 November 2006

Dirk Eddelbuettel: RQuantLib 0.2.5

The new version 0.3.14 of QuantLib was released yesterday (and I prepared Debian packages for QuantLib itself, the QuantLib-SWIG Python and Ruby wrappers, the QuantLib Reference Manual (pdf format) and the QuantLib Reference Manual (html format) last night). On the same release day, Dominick and I also updated the RQuantLib package. As a result, a new version 0.2.5 has now been uploaded to both R's master CRAN host and Debian. Only minor API updates were needed. It may take a day or two for the Debian packages to start to migrate over the mirror network. The shared library package once again introduces what is technically a new package (as the release version is encoded in the name) and we have to wait for it to be approved by the ftp masters. In the meantime, the sources are of course also here.

6 November 2006

Dirk Eddelbuettel: Golden Apple 2006

Ran the Golden Apple 5k this morning in neighbouring Forest Park. Local race, a few hundred runners, glorious weather -- what more can you ask for? I didn't feel too pumped and meant to take it easy, but still managed to beat 20 minutes at 19:44 minutes for a 6:22 min/mile pace.

27 October 2006

Dirk Eddelbuettel: Too good to be true?

Robert Cringely opens his new-look digs (now in blog format which is funny given how his columns predate the term by a good while, even though they always were web logs') with a stunning column about a new disk technology coming to market next year:
[...] disk drives that held up to three times as much data in the same space, were more reliable, actually cheaper to build, and used 70-95 percent less energy to run.
[...]
Our metal foil drive costs less, not more, and spins up so quickly that data can be read from disk as fast or faster than it can be read from flash. Who needs a hybrid disk drive?
[...]
The market potential is one billion computer disk drives and one billion mobile phone drives per year. And it all starts around this time next year when metal foil drives will begin to appear under well-known brand names.
Bring'em on.

26 October 2006

Dirk Eddelbuettel: littler 0.0.8

I just wrapped up version 0.0.8 of r (pronounced littler), put it here and sent it to Debian's incoming queue. This version has a few small changes and fixes over 0.0.7. Most notably is that all of R's environment variables are now set for r as well (at compile time). This permits it to call JGR, for example. Other than that, we made the build process more robust, extended or clarified some of the documentation and added two more example scripts.

25 October 2006

Dirk Eddelbuettel: Boston bound

This just in this morning:
111th Boston MarathonDear Dirk Eddelbuettel,

This is to notify you that your entry into the 111th Boston Marathon on Monday, April 16, 2007 has been accepted, provided that the information you submitted is accurate.

[...]
Yay! This was using the time from Sunburst in June but given Sunday's time I even get to run in 2008 too.

23 October 2006

Dirk Eddelbuettel: Chicago Marathon 2006

Today was the 29th Chicago Marathon in what the papers describe as challenging conditions: mist, some winds, and generally colder than usual. While that didn't seem to affect the elite runners and the winner (2:07:35, but a fall at the end), a lot of folks in my running group reported results that were only fine', or worse. I run fairly steady for 20 miles, then slowed a bit but did fine with a course PR of 3:18:49. That is around 7:36 min/mile and about a minute slower than the earlier Sunburst Marathon, but faster than last year or the previous year. Oh, and this also marks the end of four consecutive weeks with races.

17 October 2006

Dirk Eddelbuettel: New Quantian intro

Matthew Vona from ChemicalAnalysis.com wrote to say that they published a Quantian intro. Nice work, keep'em coming!

15 October 2006

Dirk Eddelbuettel: Makin' Tracks 2006

Ran the annual Makin' Tracks 5k this morning in support of the athletic tracks at Concordia, a collegue two blocks up from my house. As I actually train there once a week, and as the race goes straight through my neihbourhood, it's a really nice one to do. Somehow it had gotten cold the last few days, and this morning at 8:00am it was only a few degrees above freezing--yet sunny and pleasant. Pretty decent conditions after all. I managed to run pretty fast for a new PR of around 19:2159 handstopped, or just under a 6:15 min/mile pace.

14 October 2006

Dirk Eddelbuettel: Mythtv settings bitrot

Note to self, second in an ongoing series started last year. If MythTV, after an otherwise innocuous kernel upgrade complete with an upgrade of the required ivtv driver, fails to show live tv even though ivtv itself works, check the settings and make sure the second tuner on the pvr500 card doesn't point to nowhere. Still not sure how a kernel upgrade could affect the setting inside the MySQL db, but what can you say. On the other hand, the upgrade of MythTV itself from 0.19 to 0.20 was seamless. Oh well.

9 October 2006

Dirk Eddelbuettel: Frank Lloyd Wright Race 2006

The 30th annual Frank Lloyd Wright Race was today in neighbouring Oak Park. Really nice weather, once more. The race went quite well: I finished with a hand-stopped time of 41:02 (or a 6:37 min/mile pace) which is quite a bit faster than my times here in 2004 and 2003.

Dirk Eddelbuettel: littler 0.0.7

A few days ago, Jeff and I rolled up 0.0.6, but we forgot one file in 'make dist' (doh!) so version 0.0.7 of our newish toy r (aka 'littler') followed suit. In the meantime, littler was also added to Debian so you can get 0.0.7 via apt-get too.

2 October 2006

Dirk Eddelbuettel: Chicago Half Marathon 2006

Ran what may be my favourite race this morning: the Chicago Half Marathon in its tenth anniversary (having run the three previous ones too). Once again gorgeous weather, sunny, no wind, low 50s at the start. The race itself went rather well. I wasn't sure how I had weathered a bout of the flu with some light fever last weekend, but resting most of this week seems to have helped as I was able to push my PR by about 1 min 45 secs to 92:33, or a 7:039 min/mile pace.

27 September 2006

Dirk Eddelbuettel: Announcing 'littler'

Earlier today, Jeff announced our new toy r (aka 'littler') with this post on the r-help list. What is 'littler'? We summarized it briefly in the post and on our respective or pages:
GNU R, a language and environment for statistical computing and graphics, provides a wonderful system for 'programming with data' as well as interactive exploratory analysis, often involving graphs. Sometimes, however, simple scripts are desired. While R can be used in batch mode, and while so-called 'here' documents can be crafted, a long-standing need for a scripting front-end has often been expressed by the R Community. littler (pronounced 'little R' and written 'r') aims to fill this need. It can be used directly on the command-line just like, say, bc(1):
	$ echo 'cat(pi^2,"\n")'   r
	9.869604
http://ftp-master.debian.org/new.html Equivalently, commands that are to be evaluated can be given on the command-line
	$ r -e 'cat(pi^2, "\n")'
	9.869604
But unlike bc(1), GNU R has a vast number of statistical functions. For example, we can quickly compute a summary() and show a stem-and-leaf plot for file sizes in a given directory via
	$ ls -l /boot   awk '!/^total/  print $5 '   \
		 r -e 'fsizes - as.integer(readLines());
			print(summary(fsizes)); stem(fsizes)'
	   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.
	     13     512  110100  486900  768400 4735000
	Loading required package: grDevices
	  The decimal point is 6 digit(s) to the right of the  
	  0   00000000000000000011112223
	  0   5557778899
	  1   112233
	  1   5
	  2  
	  2  
	  3  
	  3  
	  4  
	  4   7		
And, last but not least, this (somewhat unwieldy) expression can be stored in a helper script:
	$ cat examples/fsizes.r
	#!/usr/bin/env r
	fsizes - as.integer(readLines())
	print(summary(fsizes))
	stem(fsizes)
(where calling /usr/bin/env is a trick from Python which allows one to forget whether r is installed in /usr/bin/r, /usr/local/bin/r, ~/bin/r, ...) A few examples are provided in the source directories examples/ and tests/.
At least one shortcoming already became apparent: by calling it r to not trample on the real R binary, we obviously overlooked that there are certain operating systems on which case is ignored -- ouch. So Mac users may have to install it as a littler binary. Other comments, as seen in the mailing list thread, are mostly supportive. littler can either be downloaded from either the Vanderbilt BioStat Wiki or my local archive. Also available is SVN access via
	svn  http://littler.googlecode.com/svn/trunk/ littler
Debian users shall get it via apt-get install littler as soon as it leaves Debian's NEW queue. In the meantime, the configure && make && make install three-step works for us under any of Debian testing, Ubuntu Dapper, OS X, and with both R 2.3.1 and the pre-releases of R 2.4.0.

20 September 2006

Dirk Eddelbuettel: EarthWallpaper.org script, anyone?

A few days ago, the Kottke blog pointed to earthwallpapers.org. While being a most excellent site, it makes uses of a newer version of flash that doesn't run on my computer. But as it happens, the flash programming is merely a shiny (and arguably fairly pretty) 'menu' to the various picture at the Flickr tag set for earthwallpapers. Better still, the author uploads daily additions, see today's in 1280x1024 pixels. Now, I don't really know much about Flickr, but didn't they open some sort of API? Is there a way write a quick Bash, Perl or Python script to have cron download the current picture as a desktop wallpaper?

8 September 2006

Dirk Eddelbuettel: More Comcast fun

One of the many recurring themes on this blog is the amount of trouble we're having with Comcast. So Ladies and Gentlemen, here is a new one: Having been forced to use the ISP as a smarthost for delivery (as reverse lookup is not provided for residential service), I depend on Comcast to deliver my mail. But look what we just got for attempting help a user on the quantlib-users list:
Reporting-MTA: dns; comcast.net
Arrival-Date: 8 Sep 2006 1:46:30 +0000

Final-Recipient: rfc822;
Action: failed
Status: 5.1.0 MAIL FROM: 550 REPLY:
550-"rejected_because_your_SMTP_server,_204.127.225.94,_is_in_the_Spamcop_RBL.
Diagnostic-Code: smtp; Permanent Failure: Other address status
Last-Attempt-Date: Fri, 8 Sep 2006 01:46:36 -0000
So a benefit of handing over around $500 per annum is that I do not get my mail delivered because the goods folks at sourceforge.net now think ATT Broadband aka Comcast [ to whom the address mail server 204.127.225.94 resolves ] is a spammer. Someone please remind me why I haven't switched to Speakeasy?

1 September 2006

Dirk Eddelbuettel: New CRAN package 'random' uploaded

A few days ago, I uploaded random to CRAN where it now has its own summary page. It has also been announced on the r-help and r-packages lists. The random packages provides convenient access to the non-deterministic random numbers provided by the random.org site created by Mads Haahr. While certain hardware and software solutions that provide access to non-deterministic random-numbers exist, few if any are portable across all the hardware platforms R supports. Retrieving non-deterministic random numbers may be beneficial to seed parallel simulations with independent draws, to obtain portable initializations for other RNGs, to validate simulation with non-deterministic RNGs, or simply for fun and experimentations. The package is fairly straightforward thanks to R's ability to read directly from remote URLs, and to random.org's nice CGI interface. Included are two pdf file ('vignettes' in R parlance). The first, random-intro, discusses the packages and provides some (currently just initial) tests using dieharder, a thorough test suite for RNGs. The second, random-essay is a 'transcriped' version of Mads essay.

16 August 2006

Dirk Eddelbuettel: RQuantLib 0.2.4

A new RQuantLib package 0.2.4 was uploaded to R's master CRAN host and Debian. The principal changes are a small ajustments to work with the new QuantLib release 0.3.13, and the addition of some first unit tests. As of right now, this version will not build in Debian testing due to problems with g++-4.1 (see #383251). It works on unstable, and should work with either a slighly older g++-4.1 or g++-4.0. However, as CRAN runs Debian, it may not be until a fixed g++ hits testing that this package will hit the CRAN mirrors. There's always my page as a fallback.

14 August 2006

Dirk Eddelbuettel: Chicago Distance Classic Half Marathon 2006

This morning, under what can only be described as unwordly gorgeous conditions, was the this year's edition of the Chicago Distance Classic. I had done this before in 2003 and 2004 but skipped it last year. In its thirtiest edition, the race has moved under the umbrella of the LaSalle Bank organisation that also produces the Shamrock Shuffle and the Chicago Marathon. In fact, it is now part of a challenge' series of three teams runs. Not sure yet how we did as a team. I was running with a friend who needed to finish in 1:35 to get a competitive start for the Marathon. And we did it -- very steady 7:15 min/mile paces made for a pretty well run race. We ran together for 4 miles, then he pulled ahead. I caught him once more, only to loose him again, but overtook him around mile 12. My chip time was 94:18 for a 7:12 pace; he came in a few seconds later. I was quite pleased how steady I ran the race; I didn't even walk a single water stop. I just checked, and this is in fact a PR for the distance. Nice. On the unpleasant side, I have the meanest blister I ever had. I had ordered new shoes from this place only to learn two weeks later that my order was lost but that I was free to issue a new order at their (higher-priced) mothership store Road Runner Sports (you know, the guys with the unstoppable supply of mail-order catalogues...). Thanks, but I didn't feel like falling for that. New order from Running Warehouse should arrive tomorrow. But I guess I need a few days to let the darn blister heal.

Next.

Previous.