Comments for tech guy in midtown

Comment on Zen and the Art of Elevators by pters

pters — Sun, 11 Dec 2011 20:21:06 +0000

At least yours gives you some sort of feedback. Ours generally works immediately for the first person, but only sometimes immediately for the second person. No flashing lights to tell you when it’s ready.

My tactic if I’m second is to hold my card over the sensor and press the button every half a second or so till it accepts it. There may be a fixed delay period I can wait and be guaranteed success, but it’s only three floors, and there isn’t time to experiment. If there is, I think it’s longer than 5 seconds.

Comment on Scatter Map of Madoff’s Victims by dimi

dimi — Fri, 15 Apr 2011 16:19:07 +0000

Hi! Nice graph! One question, how did you turn up pdf to text programmatically? That’s the most interesting….
( for the graph…now you can know where the high-income lies? more taxation? hehehehe )

Comment on Review of Programming Pearls by Jon Bentley by Ed Reingold

Ed Reingold — Tue, 09 Nov 2010 16:11:20 +0000

Guilty as charged! Here is some history behind the problem.

I thought of the problem as an exam question in an algorithms course
1972; it turned out that a student in the class (Barry Finkel, now of
Argonne Labs) told the problem to his brother, Raphael Finkel, then a
CS grad student at Stanford (now at U of Kentucky). He, in turn,
mentioned it to Jon Bentley and Bentley used it in a Programming
Pearls column in CACM (at that time a real research journal) and it
was reprinted in Bentley’s book of the same name.

My wife Ruth and I were driving from Urbana to Chicago when I thought
of the problem, and it was my wife (then a grad student in chemistry
at UIUC) who had the gist of the solution that made it such a neat
problem. (The original context was from a cynical view of the Chicago
Tribune’s “Little Fooler” crossword puzzles of the 1950s.)

I always felt bad that Ruth (we’ve now been married almost 42 years!) never got
the credit she deserved.

Comment on Efficient Calculations on Named Columns of Realworld Data by Greg

Greg — Wed, 19 May 2010 03:04:08 +0000

Hi Rob,

Yes, I’ve used PyTables — I think they’re great. I was even able to store scipy Timeseries objects in them using an extension which was in the development branch at the time That was about 2 years ago though, so it might be merged into the main PyTables by now.

Thanks for the note.

Greg

Comment on Efficient Calculations on Named Columns of Realworld Data by Rob

Rob — Tue, 18 May 2010 22:19:00 +0000

Hi Greg,

I’ve just started playing with PyTables (http://www.pytables.org/moin) which answers some of my needs for timeseries far better than a relational database… essentially it allows you to store data as numpy arrays (either fixed or extendable). Are you aware of this package? Any thoughts on its usefulness?

Regards,
Rob

Comment on Pythonic Data Analysis with MaskedArray and Timeseries by Greg

Greg — Mon, 01 Mar 2010 12:22:01 +0000

No, I hadn’t heard of Pandas. Thanks for the pointer — it looks new, interesting, and useful.

Comment on Pythonic Data Analysis with MaskedArray and Timeseries by Richard House

Richard House — Mon, 01 Mar 2010 10:33:00 +0000

Have you seen Pandas (on Googlecode)? It has some useful functionality for handling real world timeseries data, plus a lot of other quant related stuff.

Comment on About by Greg

Greg — Wed, 10 Feb 2010 23:07:49 +0000

I use Python’s pylab. In particular, I used the “boxplot” function on the axes object returned from a call to pylab.axes().

Comment on About by Joseph Reagle

Joseph Reagle — Wed, 10 Feb 2010 22:52:14 +0000

In your “Fantasy Football Pickem Strategy Results after Week 4″ entry, what did you use to print the pretty summary of stats? Your plot looks like matplotlib, but I don’t see any pretty 5-figure summary function in scipy.

Comment on Review of OmniGraphSketcher by Drazick

Drazick — Fri, 04 Dec 2009 09:43:40 +0000

Any Windows alternative?