# Some fortnights of links – 28 Sep 2013

It has been awhile since my last link round-up.  I few months ago I had gotten in the habit of book-marking items I came across online that I thought were interesting or relevant to the topics of this blog, and then posting them online every two weeks.

Jason Collins over at Evolving Economics posts a similar round-up of links weekly. A few months back, when I went to post my bi-weekly list, I discovered that he had already posted ~90% of them. So, as not to be redundant, I gave up for awhile and decided to finish my dissertation instead.

However, now that the dissertation is over, I am getting back to blogging. I also have not been reading as much on the internet lately, so this week’s list will be relatively short:

Florid or falsifiable? The use of metaphors in science. Cultured Primate’s discussion of the “ratchet effect.”

Steven Pinker embraces scientism. Bad move, I think. The internet seems to be well over Pinker’s seemingly humanities-dissing essay.  But I think Massimo Pigliucci’s response is worth a link.

Some friends may have moved to the Switzerland of Europe, but I have moved to East Tennessee, the Switzerland of America.

Journal article of the week (a new feature):

War, space, and the evolution of Old World complex societies by Peter Turchin, Tom Currie, Edward  Turner, and Sergey Gavrilets (disclosure: Gavrilets in my post-doc mentor) in PNAS.  The article is open access. It was covered in Wired and a lot of other outlets. Peter Turchin blogs about it here and here.

Entertaining in that internet sort-of-way:

Every Second on the Internet

# My New Job: Post-doctoral Fellow at NIMBioS

Today I start as one of three new post-doctoral research fellow at the National Institute for Mathematical and Biological Synthesis (NIMBioS).

NIMBioS is an NSF-funded research center hosted at the University of Tennessee, Knoxville. Most people pronounce it “nimbus” (like the raincloud). An excerpt from the NIMBioS mission:

A major goal of mathematical models and analysis in biology is to provide insight into the complexities arising from the non-linearity and hierarchical nature of biological systems. Primary goals of NIMBioS [are] to foster the maturation of cross-disciplinary approaches in mathematical biology and foster the development of a cadre of researchers who are capable of conceiving and engaging in creative and collaborative connections across disciplines to address fundamental and applied biological questions.

Read as mathematically-inclined biologists and biologically-inclined applied mathematicians doing science!

One of the best things about this post-doc is that I have the freedom to focus on my own research interests. Another is that I have an awesome research mentor. Another is that I will get to work with a great group of people. Another is that I get to bring together researchers I admire from different fields by organizing two transdisciplinary working groups.  And I get to participate in workshops already underway.

In any case, watch this space for updates about my scientific adventures in Tennessee. (Apparently, there’s a chance I’ll finally become a college football partisan.)

# Navigating Black Rock City: Geometrically-based Heurisitcs

I spent much of the past week in Black Rock City which at ~68,000 people is, for one week a year, the third largest city in Nevada.

Unlike most other cities, BRC is laid out in polar coordinates. Some roads radiate out from the center that are numbered on an hour-system (like a clock) with intervals every half-hour. Other roads follow the circumferences of the circles with the innermost road called the “Esplanade” and the rest given names following the alphabet from innermost to outermost (this year from A to L).  (See a diagram of the 2013 layout here and a huge aerial photo of the city here.)

While navigating a normal grid is pretty straight forward.  Navigating a polar grid is less intuitive for most of us. One thing that becomes clear rather quickly is that sometimes moving to an inner circle and then navigating back outward is shorter than staying on an outer circle.  Often the question arises “when is it more direct to travel inward to a smaller circle and when is it more direct to just travel around a given circumference.” This blog post is an attempt to develop a heuristic

(Note that some BRC participants may object to the very idea of trying to get somewhere efficiently.  The journey, they may argue, is often more interesting than the destination. I have some sympathy for ,this view.  However, you sometimes just want to be first in line for some cold pizza* or get the best seat for a geology lecture. Furthermore, folks who would like a higher journey to destination ratio, can simply follow the opposite heuristic from that developed in this post to extend their travel times.)

For those not interested in the mathematical details, here are the rules of thumb I derive below:

1) If your starting point is farther from the circles’ center than your destination, travel to the circumference road that corresponds to your destination (e.g. if you are starting on H and your destination is on C, travel to C).

2) If you would rather stick to the Esplanade than cut across the inner circle: If your destination is four or more hours away, take the Esplanade and then travel up the nearest radial road to your destination.  If your destination is fewer than four hours away, take the circumference road and then travel up the nearest radial road to your destination.

3) If you are willing to cut across the inner circle: Cut across the circle if your destination is 3.5 or more hours away. If you are on A or B road, cut across the inner circle if you are three or more hours away. Otherwise, stay on the circumference road and then travel up the nearest radial road to your destination.

Note that my analysis ignores the complicated business around six o’clock which is confusing enough to be avoided anyway.  And I ignore the open circle at the top.  If you are crossing it, the rule is just to take a straight line to your destination on the other side.

The Maths Part 1 (we need roads edition):

The logic behind step (1) above is pretty straight-forward.  It is shorter to travel around an inner circle than an outer circle, so if you are going to travel to a point on an inner circle, it makes sense to travel there first.  Similarly, if the point you are going to is on an outer circle from your starting point, it doesn’t make sense to move to the out circle until the end of the journey.

This insight effectively reduces the problem to one where your start and end point are on the same circle and the question becomes whether one should move to an even smaller circle that this starting point.

These two options are shown in this figure, where r is the distance from the center of the circles to the inner circle and d is the distance between the inner and outer circle:

Using the formula for the circumference of a circle (where h is the angular distance traveled around the circumference in hours) , the distance of the inner route is:

$\frac{h}{12} 2 \pi r + 2d$

The distance to the outer route is:

$\frac{h}{12} 2 \pi (r + d)$
When is the inner route shorter than the outer route:
$\frac{h}{12} 2 \pi r + 2d < \frac{h}{12} 2 \pi (r + d)$

Which reduces to:

$h > \frac{12}{\pi} \approx 3.8$

Notice that this answer is independent of both d and r.   Since the roads are at half-hour intervals, it is best to go to the Esplanade when one is four or more hours away (one third of the circumference of BRC) and it is best to stay on one’s circumference road when one is 3.5 or fewer hours away.

The Maths Part 2 (where we’re going we don’t need roads):

The above analysis assumes that one stays on the Esplanade, but it is always a shorter distance to cut across the inner circle (though the  road conditions are not as good).  Does this change the analysis?

Now the options look like this:

The length of the outer route is the same as above:

$\frac{h}{12} 2 \pi (r + d)$

However, the formula for the inner route uses the formula for the length of a chord:

$2 r \sin \left(\frac{\theta}{2}\right) + 2d$

The condition for cutting across the inner circle vs taking a circumference, where theta is in radians is:

$2r\sin \left(\frac{\theta}{2}\right) + 2d < \theta (r + d)$

Because of the sin function, solving this inequality is tricky.  So I just solved it numerically using the ratios of r to d from the official BRC layout — the distance from the Esplanade to A is 1/5 the distance from the center to the Esplanade, r, and the distance between each subsequent concentric circle is 1/10 of r.  Plugging these values in to the above inequality for various times from 0 to 6 hours gives the following conditions where it is shorter to cut across the inner circle:

 Radians 0 0.26 0.52 0.79 1.05 1.31 1.57 1.83 2.09 2.36 2.62 2.88 3.14 Hours 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 A OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN IN B OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN IN C OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN D OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN E OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN F OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN G OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN H OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN I OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN J OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN K OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN L OUT OUT OUT OUT OUT OUT OUT IN IN IN IN IN IN

Therefore, it is always shorter to cut across the inner circle when your destination is 3.5 of more hours away.  But if your starting or ending points are on A or B, it is also shorter to cross the inner circle when your destination is three or more hours away. Otherwise, it is shorter to stay on the outer circle.

* Especially when they run out of pizza by 11:55 – before their 12:00 start time.