Kindle 2 update

Amazon has provided a new firmware update that enables the Kindle 2 to read PDF natively. I have updated my K2i and it works pretty well. However, two-column papers are still hard to read because of the small screen size. Also, it seems that we cannot highlight text and/or add user comments (which is a convenient feature that I am using more and more).

Converting PDF books for the Kindle 2 in Snow Leopard

The Kindle 2 (K2i) looks like the perfect companion tool for the academic researcher. However, Amazon has decided the K2i should not be able to read PDF natively, which is kind of annoying. Each PDF file must be converted to one of the e-book formats supported by the Kindle, i.e., AZW, PRC, MOBI, TXT. Unfortunately, we do not have a good, native, 'PDF2Kindle' software for Mac OS X (converting PDF files to e-book formats is not an infallible science). Instead, the best option is to use the Mobipocket eBook creator with a Windows emulator. These are the steps you can follow to get a working system in a Macbook Pro with Snow Leopard. Also, make sure that you have installed the X11 tools from the Snow Leopard install disk. These tools are required for the normal Crossover operation.

1) Install Crossover 8.0 Professional for Mac. You can buy this software at the codeweavers web site: http://www.codeweavers.com/products/cxmac/

2) Download the windows installer (creator.msi) for the Mobipocket Creator 4.2 for Windows 2000/XP at http://www.mobipocket.com/en/downloadSoft/DownloadCreator.asp.

3) Open Crossover and create a new Windows2000 'bottle'. CrossOver allows you to maintain multiple bottles in one CrossOver installation. A 'bootle' is like having several different Windows machines operating together on your computer.

4) Select 'manage bottles' in the Crossover menu. In the newly created bottle, install the Microsoft Explorer 6 from the list of supported software packages. Follow the indications of Crossover and the software will be automatically downloaded and installed.

5) Now, we install the mobipocket creator (in the same bottle) by choosing "installing unsupported software" in Crossover. When the installer is running, select install "mobipocket publisher" (the other choice, mobipocket reader, cannot convert PDF files).

And that's all. Now, you can enjoy reasonably complex PDFs with figures in your Kindle 2. On the other hand, the above recipe does not work very well with PDF files displaying equations and other mathematical symbols (...). If do you have any means of getting scientific PDF papers on the Kindle with the Mac, please let me know!

Punto y Raya Festival 2009

This year I has been invited to participate in the "Punto y Raya" festival, which has earned the title of "most abstract in the word". This is an interesting initiative between art and science to be held at the Art Centre Santa Monica in Barcelona during November, 25-29th. Here is the abstract for my talk:

"Complex Networks: The Architecture of Life"

Scientific Studies of complex systems show that all networks - from epidemics to the electric power grid, from the cell to the internet - share similar properties, regardless of their origins. That is, what is relevant to understand their global organization is not the specific details of 'nodes and links' but how these elements are interconnected and assembled to form the complex network. Here, visualization is an indispensable tool that provides deep insights into the structure and the evolutionary and creative forces shaping complexity.

Check the festival website and the video trailer in Facebook. Update: Watch my talk (in English with Spanish subtitles) here.

Workshop: "From Insect Nest to Human Architecture" (September 23-25th, Venice)

It is now some three weeks since we have celebrated this interesting workshop in Venice (check the related website) . For me, this workshop was a kind of personal milestone: the end of a two-year post-doc in Toulouse. And I have learned a lot of things during these two years. Then, it was time for summarizing what I've done and, more importantly, what remains to be done.

A goal of the workshop was to define links between biological studies of insect nests and human architecture. However, it was difficult to draw any deep connection between them. Typical parallelisms between human and insect buildings include functional elements like thermoregulation. The bibliography cites termite mounds in the savannah as a classic example of biological thermoregulation. While it is clear that successful cultivation of fungi has specific thermal requirements (and thus imposes some structural constraints on the host termite nest), recent empirical evidence presented by Scott Turner at the workshop indicates that (at least in some cases) termite mounds do not always work as perfect thermoregulation devices. Instead, Turner discussed a very interesting hypothesis of the termite mounds as a humidity regulation device, bringing water from the soil to the mound during summer (and the other way around in winter) in order to keep the adequate level of humidity. Moreover, Turner proposed the idea the termite mount acts as a filtering device, where different parts of the mound filter different components out of environmental fluctuations. Then, we conclude there are different functional requirements between human and insect buildings.

As a complex network researcher, my own interest was to make connections between biological studies performed by field researchers (like John Menzel, Walter R. Tschinkel, Scott Turner and Flavio Roces, among others) and complex network theory. This was a natural way to proceed because I have already published some network studies on nests of Cubitermes Spp. termites (see Valverde et al, Physical Review E 79, 066106 (2009)) and we have some ongoing network works with nests coming from other species (including Trinervitermes and Procubitermes termites). Network theory enables us to look at insect nests from a fresh, new perspective. The network representation of a termite nest highlights the global pattern of interconnections instead of minute details regarding network elements, i.e., the node and the link. Indeed, we have proposed a classification method for networks based on counting how many different 4-node subgraphs occur in a network (i.e., network motifs).

In this context, network models can be quite useful to address the intrinsic limitations of empirical studies. For example, it may be quite costly to obtain enough experimental samples for a meaningful statistical study. A (valid) computer model has the potential to extend empirical studies with a few number of biological specimens to a large number of synthetic networks. This can be very valuable when testing a number of different hypothesis about the evolution and development of termite nests. Then, complex network theory has the potential to explore new biologically meaningful questions which will be difficult to answer with limited data. For example, are these networks optimal? and if affirmative, how we can measure the degree of optimality displayed by these insect nest networks? I think that we have started to answer the optimality question by comparing real insect nest networks with a simple network lattice model generated with a rule-based algorithm. In order to produce comparable structures, our algorithm is feed with the most appropriate parameters estimated directly from the real insect nest network.

On the other hand, computer renderings of networks have an aesthetic value. During my research I have produced many network visualizations of different natural and artificial systems. Many people found network renderings beautiful or even valuable from an artistic point of view. Recently, I have been asked for permission to republish a computer rendering of a large patent citation network in an incoming publication in the Art Journal (to be published, 2010). Human architecture is (and in particular, modern or contemporary architecture) related to aesthetics. Without doubt, we can find beauty in insect nests network. And it seems a perfectly valid goal to seek for inspiration in these biological structures for human architecture. Indeed, I have learned from the talk of Juhani Pallasmaa that there is a well-documented interest in biologically-inspired architecture.

However, aesthetics is a very difficult thing to measure. Still, the ancient greeks related beauty with some predefined ratios and geometric relationships. In a related research, evolutionary computation has been used to produce synthetic imagery. We can conceive a similar scheme to produce architectural designs resembling biological structures (i.e., insect nests). For example, we can create a software application that couples a human architect/designer with a genetic search algorithm that looks for the most beautiful or pleasant insect-like networks. We can conceive an hypothetical design session with this human-oriented genetic algorithm. At every step, the algorithm automatically will generate and show to the designer a small collection of random networks (e.g, and in order to produce more realistic results, we can use the computer model for insect nest networks discussed above). The designer evaluates and judges what the best design is, which in turn is used by the algorithm as a seed to create the next generation of designs. This process is repeated until the designer finds the ideal network shape, which can be used to create a physical or virtual art.

We can see that there is still a long way to go. However, we can readily appreciate the added value of using network theory when studying insect nest networks. And who knows if the same network formalism will be also useful for creative purposes, i.e., producing aesthetic, biologically-looking buildings and unique pieces of art.

Welcome to my blog

In this blog I will try to express my thoughts on complex networks, computational biology and computer graphics.
Stay tuned!
Sergi