Brian Hayes writes in the recent issue of American Scientist on An Adventure in the Nth Dimension. He beautifully explores the surprising development of the volume (and the surface) of the unit ball as the dimension increases. This is best illustrated by reproducing the graphs of the two functions (extending the domain to the positive reals). The following SAGE code (re)produces the plots which were presented by Paul Renno Heyl (Properties of the locus r = constant in space of n dimensions, 1897).

def V(n):

return pi.n()^(n/2.)/gamma(n/2.+1)

def A(n):

return n*V(n)

G = plot(A, 0, 20, color = 'red', legend_label='surface of unit ball') + list_plot([A(n) for n in srange(20)], color = 'red')

G += plot(V, 0, 20, color = 'blue', legend_label='volume of unit ball') + list_plot([V(n) for n in srange(20)], color='blue')

G.show()

Alasdair gave in his blog post Interlaced eigenvalues a small Matlab/Octave example around the following beautiful result due to Cauchy:

Let A be an n \times n Hermitian matrix with eigenvalues \lambda_1 \leq \lambda_2 \leq \dots \leq \lambda_n and let the matrix B be obtained from A by removing the ith row and the ith column for some 1 \leq i \leq n. Then the eigenvalues of B \mu_1 \leq \mu_2 \leq \dots \leq \mu_{n-1} satisfy the interlacing

\displaystyle \lambda_1 \leq \mu_1 \leq \lambda_2 \leq \mu_2 \leq \lambda_3 \leq \dots \leq \mu_{n-1} \leq \lambda_n.

Alasdair gave a little Matlab/Octave experiment to illiustrate this and we reproduce it here using SAGE (sticking to his notation for comparison).

sage: n = 4

sage: D = diagonal_matrix(range(1,n+1))

sage: D

[1 0 0 0]

[0 2 0 0]

[0 0 3 0]

[0 0 0 4]

sage: R = random_matrix(RR,n)

sage: R

[ 0.778692495501054 -0.135208587751857 0.257223918336617 0.774429743108338]

[ 0.423676053284628 0.417711760181128 0.241748509057067 0.961897138926902]

[ 0.761841048992683 0.728238125048413 -0.919597769483716 0.778538514077347]

[-0.459693778794198 -0.349046725059769 -0.969387932465504 -0.922931517520695]

sage: S = R - R.transpose()

sage: Q = (S-identity_matrix(n)).inverse()*(S+identity_matrix(n))

sage: M = Q*D*Q.transpose()

sage: M

[ 1.95237691577229 0.231093937356808 0.216951816905018 0.296266773415154]

[ 0.231093937356808 2.92188071228280 0.263114531184218 0.0747380763688735]

[ 0.216951816905018 0.263114531184218 2.90749330981640 -1.36941012136363]

[ 0.296266773415154 0.0747380763688735 -1.36941012136363 2.21824906212850]

sage: M.eigenvalues()

[3.99999999999999, 3.00000000000003, 1.99999999999998, 1.00000000000000]

sage: x = range(n)

sage: x.remove(randint(0,3)

sage: N = M.matrix_from_rows_and_columns(x,x)

sage: N

[ 1.95237691577229 0.216951816905018 0.296266773415154]

[0.216951816905018 2.90749330981640 -1.36941012136363]

[0.296266773415154 -1.36941012136363 2.21824906212850]

sage: N.eigenvalues()

[3.97504068833705, 2.09499000737375, 1.00808859200640]

Footnote: We had to take a little detour to delete some rows and columns from a matrix. It turns out, that this feature is already Ticket #11528 on the trac server of sage.

We considered patenting; we prepared a patent and it was nearly filed. Then I had an interaction with a big, multinational electronics company. I approached a guy at a conference and said, “We’ve got this patent coming up, would you be interested in sponsoring it over the years?” It’s quite expensive to keep a patent alive for 20 years. The guy told me, “We are looking at graphene, and it might have a future in the long term. If after ten years we find it’s really as good as it promises, we will put a hundred patent lawyers on it to write a hundred patents a day, and you will spend the rest of your life, and the gross domestic product of your little island, suing us.” That’s a direct quote.

Source: Andre Geim: in praise of graphene, interview with nature news, published online 7 October 2010, doi:10.1038/news.2010.525 (last accessed: 9 Oct 2010)

 

 

 

npr, 5 Worries Parents Should Drop, And 5 They Shouldn’t

Based on surveys Barnes collected, the top five worries of parents are, in order:

  1. Kidnapping
  2. School snipers
  3. Terrorists
  4. Dangerous strangers
  5. Drugs

But how do children really get hurt or killed?

  1. Car accidents
  2. Homicide (usually committed by a person who knows the child, not a stranger)
  3. Abuse
  4. Suicide
  5. Drowning

Terry Tao, A first draft of a non-technical article on universality on universality and emergence (though he does not explicitely name the latter).

Donald E. Knuth, Letter to Alison J Brimelow, President, European Patent Office, 24 April 2009. Attachments include a letter of his to the US Patent Commissioner (1994) and the transcript of a talk by him at the TU Munich (2001) with related information.

Switzer et al, Operating system shut down, 31 August 2010.

Luca Lucarini, Patent Absurdity: how software patents broke the system, 2010.

Sebastian Krämer, Gemein-Freiheit, SWR 2 Wissen.

Erwin Wagenhofer, We feed the world, 2005.

FAO estimates that 1.02 billion people are undernourished worldwide in 2009.

FAO, the state of food insecurity in the world, 2009, p. 2.

Follow

Get every new post delivered to your Inbox.