Month: March 2009

Went to Queen Anne to code yesterday.  Queen Anne only because it is a neighborhood in Seattle I usually try to avoid.

All locally-owned coffee shops were too busy and Starbucks had several tables adjacent electrical outlets.  Taking one for the team–the team being me and my laptop–I stayed.

[Statement here about clientele exactly fitting my constructed Starbucks-customer stereotype (i.e. self-absorbed, overpaid, “professionals”) and how background music was selected to be noticeable but not offending to the privileged-liberal ear.]

Setting up at a table beside a window beside the sidewalk beside an important Queen Anne arterial, a glint of light strafed my eye.  Tiny blue cubes bounced on the sidewalk outside.  Looking up I see the lower half of a man’s body, his upper half bent over inside the newly-formed passenger side cavity of a late model SUV.  A half second later he emerged and jumped into the backseat of a blue Volvo waiting in the next parking space.  It sped off, traffic miraculously absent.

The vehicle’s owner, Starbucks-frequenter-extraordinaire, materialized from the cafe, vanilla latte in hand, let out a whoop, and promptly called the cops (via an iPhone of some random gawker-samaritan happening by).  Two cops, complete with semi-automatic handguns, billy clubs, mace and many other beltable tools, arrived on their nifty bikes.

Cops and female victim swept up the autoglass into a 2 quart cardboard Starbucks Coffee-On-the-Go® container.  Statements were made and taken.  Eavesdropping revealed the woman’s purse was stolen though I could see her fat card-containing wallet in hand; she obviously had to pay for her latte.

She drove off, probably in search of an autoglass shop.  The whole episode probably not much more than an inconvenience for her I’m sure.

If anyone has disdain for “sanctity of property”, and other bullshit along those lines, it is I.  Still, those crooks were seasoned thieves.  They could be using their able bodies to work, not rippin’-and-runnin’.  Pissed me off it did.

Perhaps middle age conservatism is starting to set in.

“Whoa, have we met before?  You look familiar…” I often get from people on meeting them.

Until a week ago, I figured mine was just one of those faces.

Driving home on Lake City Way last week I saw this:

[Actually, the billboard depicted here resides in Massachusetts.  Stole this pic I did from the live journal page: Run humans run which, incidentally, I recommend you check out.]

If I were to guess how I’d look in 30 years, this would be it.  Of course, I doubt I will look so jolly nor do I expect to be practicing “the most stultifyingly oppressive brand of Christianity ever known to man.”  America (The Book) pg. 18, 2004 Warner Books.

Recently, my mom told me that I cannot bear to leave something un-understood if I, myself, understand that thing.  Introspection never revealed to me this tendency so I dismissed her observation as a product of mother-bias.

Frustrated, frazzled and frantic, a student emailed me for help on his final project that he must present tomorrow.  I agreed to help him; it is my job as TA after all.  Back and forth we went for two and a half hours this afternoon, I not understanding his project any better than he.  Invertebrate phototransduction cascades and Jack-shit were identical phenomena to my knowledge.

I persevered out of pride.  Precious few math/physics problems have defeated me and this PoS I would not add to that list.

Using a basic enzyme-substrate reaction detailed in one of my textbooks, I deduced how differential equations can be derived from chemical ones so as to model reaction kinetics.  We then used this recipe to write the specific set of diff. eqs. for the student’s transduction diagram/nightmare.  The model was simplified–several nonessential intermediate reactions we ignored out of not-giving-a-shit–but all core players were included.

Vindicated we felt when the model generated expected phototransduction kinetics coarsely, but to reasonable accuracy.  I spent 20 minutes thoroughly explaining the equations, and how they were derived, so the student could further explore and fine-tune the model tonight (and he had better).

He left happy and satisfied.  So did I.

If my research gig doesn’t pan out, I guess I can make myself useful teaching. 

What follows are quotes I’ve extracted from the March 23, 2008 New York Times article Depression, You Say? Check Those Safety Nets, by Charles Duhigg. These I have categorized under “Then”. Quotes concerning the same issues “Now” I have selected from yesterday’s New York Times opinion article This Is Not a Test. This Is Not a Test. by op-ed columnist Thomas L. Friedman.

Then:

Now:

Then:

Now:

Then:

Now:

Second, we need to get a market going that would bring fair value and clarity to the “toxic mortgages” crippling the balance sheets of our major banks. This will likely require some degree of government subsidy to private equity groups and hedge funds…

Unfortunately, the president may have to look the American people in the eye and explain that “fairness is not on the menu anymore.” All that’s on the menu now is whether or not we avoid a system meltdown — and this will require rewarding some new investors.

Third, the president may have to make some trillion-dollar decisions — like nationalizing major banks or doubling the economic stimulus — with no real precedent and without knowing all the long-term ramifications.

I wish earned my pay and prestige wrongly predicting outcomes of a complicated multi-variate system that nobody fully understands. My windbaggery is on par with these two.

 

PS:  If you’re thinking of hacking into my bank account, go for it.  It won’t be worth your effort!

This might take a couple minutes to load…apologies.

Wracking my brain for something semi-interesting to post this week—anything really—I decided to model the baseball metaphor of sexual intimacy using a Markovian random walk algorithm. Since all I do is work these days [whine], work-related subject matter is all I’ve got to write about (Markov models being work-related, not sexual intimacy).

According to Wikipedia (http://en.wikipedia.org/wiki/Baseball_metaphors_for_sex) what follows are broadly accepted descriptions of each metaphorical base:

First base:…………………………………………………….mouth-to-mouth kissing
Second base:…………………………………………………groping underneath shirt
Third base:……………………………………………………mutual masturbation
Home (after rounding the bases):…………………….sexual intercourse

The Model. Dude has a 50/50 chance of making it to the next base on any given “play”. If he succeeds, he either continues to next base in the sequence—where he must then make another play—or he “scores” if his successful play is from 3rd. If a play toward any of the bases fails, he’s out, meaning: he must “bat” again with some new chick and attempt to round the bases starting from scratch. The matrix below captures this scenario in its entirety. It operates on dude’s current state-vector (representing the base he “occupies”) to determine his probability of making the next base in the sequence.

Transition probability matrix (“LayMatrix” in routine below):

Results. Based on this model, the half-successful dude should expect to make exactly TWENTY-NINE total plays to score. In the course of these 29 plays, he fails or “outs” with an average of 15 girls on the way to scoring with one.

Discussion. You’re probably saying to yourself: “Thug, was high school prom the last time you “scored”? Chance probability of success for each play is not realistic.” For those of us single in our third-decade-of-life-and-then-some, a more realistic progression matrix might be:

For this model (LayMatrix2), dude need only commit an expected 19-20 plays to attain score-status; though he should expect to fail with a chance-comparable 14 girls in the process. In this scenario, most failures occur trying to reach first base.

Comparing outcomes predicted by the two models, dude should score with fewer plays than chance under the “more realistic” model but will strike out with almost the same number of chicks. These models assume chicks’ acceptances of dudes’ advances obey Markovian point-process statistics (i.e. chick will accept or reject a dude’s play independent of his record of previous plays). My limited experience agrees. How else could so many ass-douches strut around with such fine wool.

Covering my ass. For those in doubt:

z0 = [1, 0, 0, 0, 0]’; % Initial state vector: at bat at home plate

LayMatrix = [0.5, 0.5, 0, 0, 0; 0.5, 0, 0.5, 0, 0; 0.5, 0, 0, 0.5, 0; 0.5, 0, 0, 0, 0.5; 0, 0, 0, 0, 1];
LayMatrix2 = [0.8, 0.2, 0, 0, 0; 0.5, 0, 0.5, 0, 0; 0.3, 0, 0, 0.7, 0; 0.05, 0, 0, 0, 0.95; 0, 0, 0, 0, 1];

ScoreProgressionStructAge18 = CalcExpectedNumTrialsToLay(LayMatrix,z0,500)
ScoreProgressionStructAge31 = CalcExpectedNumTrialsToLay(LayMatrix2,z0,500)

function OutputStruct = CalcExpectedNumTrialsToLay(LayTransProbMatrix,s0,nCeiling)

CumProbVecOfScoring = NaN*ones([1,nCeiling]);
CumProbVecOfFailing = NaN*ones([1,nCeiling]);
nVec = 1:(nCeiling-1);

T = LayTransProbMatrix’;

[eigVecs eigVals] = eig(T);

for i = 1:nCeiling

sNext = real(eigVecs*(eigVals.^i)*inv(eigVecs))*s0;
CumProbVecOfScoring(i) = sNext(end);
CumProbVecOfFailing(i) = 1-sNext(1);

end

ExpectedNumTrialsToScore = sum(nVec.*diff(CumProbVecOfScoring));
ExpectedNumTrialsToFail = sum(nVec.*diff(CumProbVecOfFailing));

OutputStruct.CumProbVecOfScoring = CumProbVecOfScoring;
OutputStruct.CumProbVecOfFailing = CumProbVecOfFailing;
OutputStruct.ExpectedNumTrialsToScore = ExpectedNumTrialsToScore;
OutputStruct.ExpectedNumTrialsToFail = ExpectedNumTrialsToFail;

end

ERRATUM: Expected number of girls with whom dude should expect to fail on way to scoring should be reduced by 1 for both models (i.e. from 15 to 14 in the “chance” model and from 14 to 13 in the “realistic”). True to ‘tard-dom, yours truly included dude’s success-girl in each of the above failure counts.