Continue reading »]]> Jason Dykes showed this at the GeoVisualisation Meeting in Oxford on August 30^th: A map of road crashes in Great Britain (not the UK) from 1999 to 2010. It mirrors of course the road network and the density of population but it is a remarkable graphic.

It was made by the BBC and first published on 2^nd December 2011. Apologies to those who have seen it but it resonates with quite a lot of stuff on the SIMULACRA blog including Jon Reades and Joan Serras visualisations of network traffic and flows in London and GB. There is a nice movie of all this and you can zoom into different parts of the Great Britain Map.

Continue reading »]]> After giving up on Gephi (again, I really should learn), I decided it was time to get to grips with Python and iGraph since I really need to produce multiple iterations of a graph. The matmos at CASA have, of course, been touting Python for ages, but I’ve just not had the time/incentive to install and, more importantly, actually get around to using it… until now. Of course, like all such migrations, there has been a bit of a learning curve, so I’m documenting this to save you (and me) the trouble later because it turns out that Lion and Mountain Lion aren’t 100% iGraph friendly out of the box.

iGraph itself is easy to install, you can just download and run the installer from SourceForge but the supporting libraries to enable plotting have some issues. Below is what I adapted from information provided by the people at  ffMpeg.

You will need to start by grabbing xz, which is (yet another) compression/decompression library promising better performance than the venerable gz or bzip:

./configure
make && sudo make install


You need xz so that you can decompress glib, but before you can compile glib, you will also need the latest version of gettext. Furthermore, per the instructions you will need to manually edit one of the files to get it to compile on Lion or Mountain Lion. Namely, open:

gettext-tools/gnulib-lib/stpncpy.c

and change it such that you have added a line reading:

#undef stpncpy

just before the line reading:

#ifndef weak_alias

From there, it’s your usual:

./configure
make && sudo make install


Now you can finally move on to installing glib, which will be need for compiling Cairo properly. Note that this is the point where you have to start passing in more and more command line arguments to configure:

xz -d glib-2.33.10.tar.xz
tar -xvf glib-2.33.10.tar
cd glib-2.33.10
LIBFFI_CFLAGS=-I/usr/include/ffi LIBFFI_LIBS=-lffi ./configure; make && sudo make install


Phew, we’re getting there. Now it’s time for pkg-config:

GLIB_CFLAGS="-I/usr/local/include/glib-2.0 -I/usr/local/lib/glib-2.0/include" GLIB_LIBS="-lglib-2.0 -lgio-2.0" ./configure --with-pc-path="/usr/X11/lib/pkgconfig:/usr/X11/share/pkgconfig:/usr/lib/pkgconfig:/usr/local/lib/pkgconfig"
make && sudo make install


I was able to skip straight to installing pycairo because I had already installed Cairo as part of my last update to QGIS. The excellent KyngChaos has a handy set of installers to keep things simple.

And finally it’s time for pyCairo (you probably want py2cairo since OSX is using Python 2.7):

./waf configure
./waf build
sudo ./waf install

You may also care to add this line to your

.bash_login

file:

export PYTHONPATH="/usr/lib/python2.7:$PYTHONPATH"

(You may also not want to add python2.6 or python2.5 to your Python path since this will give you some interesting errors… as I discovered.)

Continue reading »]]>

An interesting movie is currently trending visualising the last thousand years of European history and conflict. Europe resembles a large game of Risk or Civilization. Such games typically end with a single large empire, whilst the dynamics of European history show the larger empires forming only fleetingly, soon to be dissolved by the re-emergence of persistent historic nations and peoples.

I’ll resist drawing any analogies with our current European crisis, only to say that after a millennia of conflict, we must be grateful for living in a period of relative peace!

Continue reading »]]> As I get to better grips with the full richness of the Oyster data set and the complexity of the TfL network it’s gradually getting easier to build better visualisations. One of the ones that I’ve wanted to revisit for quite some time was my original ‘pulse of the city’ animation (you can see it here).

I’ll discuss some of the issues encountered in creating this video at a later date, so for now just enjoy:

Continue reading »]]> I’ve been meaning to post this for ages but have had a great deal on my plate (more posts and visualisations to follow in the next week I hope) so this has kept slipping, together with the six or seven other ‘draft’ posts I’ve got going.

Anyway, this visualisation shows average entries at each and every Underground, Overground, and DLR station over the course of a week using a 10-minute interval. So in theory there are some 300 * 7 * 24 * 6 data points in this image. Or 43,200 for those of you who, like me, are having difficulty with the mental arithmetic. Of course, I’ve deliberately gone for an artistic angle to this image so you’ll find nary a scale barnor a station label, but I hope that you enjoy puzzling out the approximate ordering within each group of lines.

The rest of the text (which is too fuzzy in the raster format) reads:

The poster shows every entry to each of London’s 330 Underground, Overground and Docklands Light Railway stations over the course of a week. Stations are coloured according to the lines that they serve. A station that serves more than one line will appear more than once on the poster. During rush hour, there may be more than 8,000 entries to a single station in just 10 minutes!

For those who are interested in the mechanics of how it was made: I pulled the aggregate data from a MySQL database, generated the base graph in Python (a big ‘thank you’ to @FryRSquared for that), and then resized the chart and added the text in Illustrator. So, with the exception of that last it’s not a bad day for Free/OS software!

Also, clearly a big ‘thank you’ too to Transport for London and, in particular, the Oyster card team who are making all of this fun work possible.

We hope to have two poster-sized versions available via print-on-demand in the not-too-distant future, so stay tuned!

Continue reading »]]>  

Full Details of the Meeting are Here

Various people from UCL and Kings are contributing to this meeting in Berlin. Mike Batty from CASA is talking on how cities and their evident complexity require big data which is rapidly becoming available, Phil Treleavan from CS at UCL is talking about experimental computational finance in a bid data environment, Tiziana de Matteo from Kings is talking about embedding high dimensional data on networks, and there are many other talks, including Gene Stanley on interdependent networks and switching phenomena. Gene was the first visitor we ever had at CASA and he came with Hernan Makse in January 1996 when I (Mike) was the only employee and we were working still on DLA models. We wrote a paper on this which was in Physical Review E in 1999 and as Gene had written with George Weiss who had written with Joseph Gillis who in turn had co-authored a paper with Paul Erdos, that makes my (Mike’s) Erdos Number 4!! But tens of thousands of people have this number. It’s a small world after all.

Continue reading »]]> The Royal Netherlands Academy of Arts and Sciences (KNAW) and the Netherlands Organisation for Scientific Research (NWO) are organising this seminar on 5^th March 2012 starting at 9-30am at Trippenhuis, Kloveniersburgwal 29, the Netherlands.

Mike Batty is talking about Complexity in Cities: Are Cities Becoming More and More Complex?, Martin Novak is talking about Evolving Cooperation,  Dan Braha about Physical Complexity in Socio-Economic Systems, Kees Stam about Crises in the Brain. What Can We Learn from Modern Network Theory?

Continue reading »]]> Here is a  new book on complexity and cities entitled “Complexity Theories of Cities Have Come of Age: An Overview with Implications to Urban Planning and Design” edited by Juval Portugali, Han Meyer, Egbert Stolk and Ekim Tan with the intriguing title that what we do has come of age. Well maybe, maybe not, I leave you to be the judge of that. But it does represent a sea change. The book has a wide cast of authors and the focus is on implications for urban planning and design. My own contribution written with Stephen Marshall reviews the origins of the field, returning to Geddes, Jacobs and Alexander, and is entitled: The Origins of Complexity Theory in Cities and Planning”. Amongst those contributing are Hermann Halken, Peter Allen, Nikos Salingaros, Bill Hillier, Jeff Johnson, Hans Meyer, Egbert Stolk, Ekim Tan, Denise Pumain, Harry Timmermans, Stephen Read, Ward Rauws, Carl Gershensen, Dirk Sijmons, Theodore Zamenopolous, Katerina Alexiou, Michael Bitterman, Sevil Sarijildiz, Ozer Ciftcioglu, and of course Juval Portugali. Hope I haven’t left anyone out.

Juval Portugalihas also written a fascinating book called Complexity, Cognition and the City, recently published too by Springer and this is also essential reading. Books are coming thick and fast on the complexity viewpoint as it is being applied to cities. There is a sense that these foundations will last somewhat longer than the earlier attempts 40-50 years ago involving a Systems Theory of Cities, largely because complexity theory is a much broader church, with wide implications for how we approach the development of ideas, innovations and knowledge in post-industrial society.

Continue reading »]]> Martin was an RA in CASA and then lecturer from 1996 to 2004. He is now Senior Lecturer in Geography at the University of Manchester. He coauthored the book with another friend of CASA, ROb Kitchin of NUI. Here is the citation

On behalf of the Association of American Geographers, …, Code/Space: Software and Everyday Life, published by MIT Press, has been selected to receive the 2011 AAG Meridian Book Award for the Outstanding Scholarly Work in Geography.  This distinction recognizes one book published last year that makes an unusually important contribution to advancing the science and art of geography.

 In selecting your book, the committee wrote:  “We feel it pushed the envelope as it explained the linkages between software and human behavior in a spatial context.  This book articulates how space and software have become so intertwined that they constitute one another in our lives. It is one of the rare books that link critical social theories with technology and philosophy. Using everyday spaces, it demonstrates how such spaces are transformed by code and how new spaces of interactions are recreated. It is the type of book that can interface with many different disciplines. It is one of the few geography books taking the technology and the potential in reconstituting space seriously.”

Continue reading »]]> Satellite Meeting: Complexity in Spatial Dynamics (COSMIC)

Location: Brussels, 5^th of September 2012

Organizers: Peter Nijkamp, Michael Batty, Stewart Fotheringham, and Emmanouil Tranos

Background

New bottom up, digital data collected for entire populations, has started being utilised in urban science and geospatial analysis for understanding and modelling urban dynamic processes. Such data will provide dramatically new insights into urban change which manifest themselves in often discontinuous forms which can be articulated using a variety of reaction-diffusion dynamics incorporating catastrophe, chaos, bifurcations, and phase transitions.

Despite these developments, urban analysis is still lacking a typology of urban dynamic processes such as urban flow data, the associated networks and interactions from labour markets to pedestrian movement, to guide the development of models using new digital data collected in real time from electronic transactions such as phone lines, electronic ticketing, and related geosensing.

Urban analysis needs to be supported by the development of theories and models that explain, simulate and predict the dynamics of cities defined across spatial/geographical scales from the global to the local, from the world city to the village. Research in this area has immediate applicability to city planning, urban policy analysis, urbanisation, and world development. In the last twenty years, the field has embraced new developments in complexity theory based on the inescapable logic that such systems mainly develop organically, from the bottom up, illustrating fascinating, surprising and sometime chaotic patterns of emergence, which show order at all scales and are hard to understand as anything but the remorseless action of decision-making at the lowest levels. This presents one of the grandest of challenges to policy-making: current policy instruments are often pitched at the wrong scale, producing methods of intervention which are largely ineffective in that they ignore the essential logic of the way such human systems actually develop.

What makes this particularly opportune is the fact that massive new streams of data with respect to movement and location patterns in city systems are rapidly becoming available. These are providing the momentum for new developments in theory and modelling and they are essential in taking complexity science to the point where it becomes truly applicable in policy making.

Submitting papers

Methodological, empirical and conceptual papers demonstrating how long-standing ideas about urban dynamics can be tested and validated using new data sources that provide information about routine decision-making concerning locations and interactions are welcome. We are particularly interested in models of the mechanisms governing how cities change, which generate both smooth and abrupt change reminiscent of criticality, catastrophe, and chaos, and can be tested and extended using these new digital data. One of the key issues in this new view of urban dynamics is the recognition that dynamics exists on all temporal scales from fast and routine, to slow and infrequent, permeated by continuous secular changes as massive discontinuities. We welcome papers which examine the dynamics of change in urban networks, their clusters, communities and hubs. Submission now open! All participants at the satellite are also expected to register for ECCS12.

Organizational details

The COSMIC Satellite Event will be a one-day event. The list of invited speakers includes among others Mike Batty, Peter Nijkamp, Alan Wilson and Stewart Fotheringham. Younger researchers will also be encouraged to submit papers. Authors are invited to submit a one-page abstract through the conference website.