Monthly Archives: April 2008

Google StreetView for Apartment/House Hunting

We are trying to find a place to live in St.Paul, while living in New Orleans and Baltimore. This Friday we will actually be going to St.Paul to look at places, but we need to have a good list of places to look at when we arrive. To that end, we’ve been using some online services, combined with a realtor.

The advent and recent expansion of Google StreetView has changed this process dramatically. Whereas before we were limited to seeing street layout and satellite images in Google Maps with Housing Maps (a mashup with CraigsList), it is now possible to take an address from an ad and go for a virtual stroll through the neighborhood.

It’s a lot easer to get a feel for a neighborhood by looking at storefronts, cars, intersections, parks, and yes, even people out and about, than it is to do so by map or satellite. Some people are up in arms about StreetView, and not without good reason, but the benefits are pretty tantalizing. People are antsy about being caught on camera, but I’ve only seen 10 or so people using the service. Most of the time residential streets are empty during the day — everyone’s at work and school. The people that I have seen are generally just walking down the street or riding bicycles. However, I did see some people with police cars parked out front of their house, apparently discussing something with officers and several family members. I couldn’t tell if it was a burglary report, a domestic violence complaint, or murder investigation, but those people probably didn’t want the whole world watching.

Have you used StreetView?

What would have happened?

Simulations of real systems make it possible to test “what if” questions, and compare the results with what did happen. For example, I am working on a study in which a stimulus pulse turns on and off while a spiral wave rotates about an obstacle. We can ask, “What happened to the spiral wave following application of a stimulus pulse?” The answer depends on many factors. The image below shows the results of an experiment that tests this question:
095_08_pulse_only
It’s hard to say what happened to the wave. We see recovery, and then eventually activation. We might expect that the additional current supplied by the pulse (green, units not shown) would lead to an earlier activation. How can we say? What if we re-run the same simulation, but do not apply a stimulus pulse?
095_08_both
On inspection, pulse application postponed activation (in the red trace) by extending the duration of the wave and thus, the refractory period. Without the “what-if” data (black) for comparison, the effect of the stimulus pulse seemed less clear.

I give here only one example, but this concept applies to simulations in general. Deterministic simulations have the advantage of perfect repeatability, something much sought after but difficult to achieve in experimental studies.

I CAN HAZ PROGRAMMING LANGUAGE?

Okay, this is pretty much just for the programmers out there that are familiar with the LOLCATS phenomenon, but there’s an esoteric programming language called LOLCODE.

It made me lol for real. Here’s an example that prints the date followed by the beginning of the Fibonacci sequence:


IN MAI datetime GIMME date LIKE DATE

SO IM LIKE FIBBING WIT N OK?
    LOL ITERATE FIBONACCI TERMS LESS THAN N /LOL
    SO GOOD N BIG LIKE EASTERBUNNY
    BTW, FIBONACCI LIKE BUNNIES! LOL
    U BORROW CHEEZBURGER
    U BORROW CHEEZBURGER
    I CAN HAZ CHEEZBURGER
    HE CAN HAZ CHEEZBURGER
    WHILE I CUTE?
        I AND HE CAN HAZ HE AND I ALONG WITH HE
        IZ HE BIG LIKE N?
            KTHXBYE
        U BORROW HE

IZ __name__ KINDA LIKE "__main__"?
    COMPLAIN "NOW IZ" AND DATE OWN today THING
    IZ BIGNESS ARGZ OK KINDA LIKE 1?
        N CAN HAS 100
    NOPE?
        N CAN HAS NUMBR ARGZ LOOK AT 1!!
    GIMME EACH I IN UR FIBBING WIT N OK?
        VISIBLE I

I think that basically speaks for itself. You can find the Python-based interpreter here. Srsly.