Big Data Meets the Eye

by Sabine Hossenfelder Physics

Remember when a 20kB image took a minute to load? Back then, when dinosaurs were roaming the earth?

Data has become big.

Today we have more data than ever before, more data in fact than we know how to analyze or even handle. Big data is a big topic. Big data changes the way we do science and the way we think about science. Big data even led Chris Anderson to declare the End of Theory:

“We can stop looking for models. We can analyze the data without hypotheses about what it might show. We can throw the numbers into the biggest computing clusters the world has ever seen and let statistical algorithms find patterns where science cannot.”

That was 5 years ago. Theory hasn’t ended yet and it’s unlikely to end anytime soon. Because there is slight problem with Anderson’s vision: One still needs the algorithm that is able to find patterns. And for that algorithm, one needs to know what one is looking for to begin with. But pattern finding algorithms for big data are difficult. One could say they are a science in themselves, so theory better not ends before having found them.

Those of us working on the phenomenology of quantum gravity would be happy if we had data at all, so I can’t say the big data problem is big on my mind, but I have a story to tell. Alexander Balatsky recently took on a professorship in condensed matter physics at Nordita, and he told me about a previous work of his that illustrates the challenge of big data in physics. It comes with an interesting lesson.

Electron conducting bands in crystals are impossible to calculate analytically except for very simplified approximations. Determining the behavior of electrons in crystals to high accuracy requires three-dimensional many-body calculations of multiple bands and their interactions. It produces a lot of data. Big data.

You can find and download some of that data in the 3D Fermi Surface Database. Let me just show you a random example example of Fermi surfaces, this one being for a gold-indium lattice:

The Fermi-surface roughly speaking tells you how electrons are packed. Pretty in a nerdy way, but what is the relevant information here?

The particular type of crystal Alexander and his collaborators, Hari Dahal and Athanasios Chantis, were interested in are so-called non-centrosymmetric crystals which have a relativistic spin-splitting of the conducting bands. This type of crystal symmetry exists in certain types of semiconductors and metals and plays a role in unconventional superconductivity that is still a theoretical challenge. Understanding the behavior of electrons in these crystals may hold the key to the production of novel materials.

The many-body, many-bands numerical simulation of the crystals produces a lot of numbers. You pipe them into a file, but now what? What really is it that you are looking for? What is relevant for the superconducting properties of the material? What pattern finding algorithm do you apply?

Let’s see...

Human eyes are remarkable pattern search algorithms. Image Source.

The human eye, and its software in the visual cortex, is remarkably good in finding patterns, so good in fact it frequently finds patterns where none exist. And so the big data algorithm is to visualize the data and let humans scrutinize it, giving them the possibility to interact with the data while studying it. This interaction might mean selecting different parameters, different axes, rotating in several dimensions, changing colors or markers, zooming in and out. The hardware for this visualization was provided by the Los Almos-Sandia Center for Integrated Nanotechnologies, VIZ@CINT; the software is called ParaView and shareware. Here, big data meets theory again.

Intrigued about how this works in practice, I talked to Hari and Athanasios the other day. Athanasios recalls:

“I was looking at the data before in conventional ways, [producing 2-dimensional cuts in the parameter space], and missed it. But in the 3-d visualization I immediately saw it. It took like 5 minutes. I looked at it and thought “Wow”. To see this in conventional ways, even if I had known what to look for, I would have had to do hundreds of plots.”

The irony being that I had no idea what he was talking about. Because all I had to look at was a (crappy print of) a 2-dimensional projection. “Yes,” Athanasios says, “It’s in the nature of the problem. It cannot be translated into paper.”

So I’ll give it a try, but don’t be disappointed if you don’t see too much in the image because that’s the reason d’être for interactive data visualization software.

3-d bandstructure of GaAs. Image credits: Athanasios Chantis.

The two horizontal axis in the figure show the momentum space of the electrons into the directions away from the high symmetry direction of the crystal. It has a periodic symmetry, so you’re actually seeing four times the same patch, and in the atomic lattice this pattern goes on to repeat. In the vertical direction, there are two different functions shown simultaneously. One is depicted with the height profile whose color code you see on the left and shows the energy of the electrons. The other function shown (rescaled) in the colored bullets, is the spin-splitting of three different conduction bands; you see them in (bright) red, white and pink. Towards the middle of the front, note the white band getting close to the pink one. They don’t cross, but instead they seem to repel and move apart again. This is called an anti-crossing.

The relevant feature in the data, the one that’s hard if not impossible to see in two dimensional projections, is that the energy peaks coincide with the location of these anti-crossings. This property of the conducting bands, caused by the spin-splitting in this type of non-centrosymmetric crystals, affects how electrons travel through the crystal, and in particular it affects how electrons can form pairs. Because of this, materials with an atomic lattice of this symmetry (or rather, absence of symmetry) should be unconventional superconductors. This theoretical prediction has meanwhile been tested experimentally by two independent groups. Both groups observed signs of unconventional pairing, confirming at a strong connection between noncentrosymmetry and unconventional superconductivity.

This isn’t the only dataset that Hari studied by way of interactive visualization, and not the only case where it wasn’t only helpful but necessary to extract scientific information. Another example is this analysis of a data set from the composition of the tip of a scanning tunnel microscope, as well as a few other projects he has worked on.

And so it looks to me that, at least for now, the best pattern-finding algorithm for these big data sets is the eye of a trained theoretical physicist. News about the death of theory, it seems, have been greatly exaggerated.

Write a comment

Karl-Heinz Rädler Awarded the Schwarzschild Medal 2013

by Axel Brandenburg Awards

During their meeting on November 19th, members of the Astrophysics Group celebrated the award of this years Schwarzschild Medal to Karl-Heinz Rädler from the Leibniz-Institute for Astrophysics in Potsdam who is one of Nordita's frequent visitors. This medal is awarded annually by the German Astronomical Society to astronomers of highest scientific standing.

Karl-Heinz is one of the fathers of modern mean-field dynamo theory. In seminal papers between 1966 and 1969, he identified several effects that characterize the evolution of large-scale magnetic fields in cosmic bodies such as the Sun and other stars, planets, and even galaxies.

During the 1970s, he developed numerical approaches to compute global nonaxisymmetric mean-field models. Toward the end of the 1980s, he was allowed to travel from East Germany to the West and spent extended periods also in Helsinki, where he computed numerically the first nonlinear, nonaxisymmetric dynamo solutions that were relevant for explaining the magnetic fields in rapidly rotating stars. During the 1990s, acting as the founding director of the Astrophysical Institute Potsdam (now the Leibniz-Institute) he participated in preparatory calculations for the Karlsruhe dynamo experiment that demonstrated self-excited magnetic fields in helically flowing liquid sodium. In recent years, he developed a method for computing numerically turbulent transport coefficients from simulations. Much of his work has been instrumental for recent developments in the astrophysics group at Nordita.

During his last visit he computed analytic solutions to certain spatially periodic flow patterns called Roberts flows that we now understand as producing highly unconventional types of large-scale dynamos. In the picture, he explains striking properties of such flows that will be part of a new paper he is currently writing with members of the group at Nordita.

Members and visitors of the Nordita Astrophysics group, left to right: Bidya Binay Karak, Ebru Devlen, Anthony van Eysden, Sarah Jabbari, Nishant Singh, Matthias Rheinhardt, Illa R. Losada, Mikhail Modestov, Karl-Heinz Rädler, Elizabeth Cole, Dhruba Mitra, Lars Mattsson, Mikhail Liberman, and behind the camera Axel Brandenburg.

Write a comment

New Arrivals to Nordita

by Hans Mühlen Staff

Postdoctoral Fellows

Nordita Fellow Lars Mattsson
Astrophysics

Lars received his PhD in 2009 from Uppsala University, under the supervision of Susanne Höfner, and later (in 2010) he joined the Dark Cosmology Centre at the Niels Bohr Institute, where he stayed for three years working in close collaboration with Anja C. Andersen. His main scientific interest is cosmic dust and his previous research has largely been about simulating dust formation together with dust-driven winds from carbon stars, and various aspects of the build-up of dust and related key-elements elements in galaxies. He joined the astrophysics group at Nordita in November 2013 as a postdoctoral fellow (Nordita fellow) and will work on physical modelling of cosmic-dust processing.

Write a comment

Solar Physics Research at Nordita

by Sabine Hossenfelder Outreach, Physics

The latest of our videos about research and researchers at Nordita portrays our astrophysics group. Learn about the students' and postdocs' daily research life and hear Prof Axel Brandenburg explain some of the group's recent findings, solar weather and the formation of sun spots. Part of the movie was shot at the recent public outreach event "Physics in Kungsträdgården."

For more details about astrophysics at Nordita, see also Oliver's video and our recent feature article about the Sun's Butterfly Diagram.

For other video presentations of research at Nordita, see   www.nordita.org/video

Write a comment

From String Theory to Strange Metals

by Sabine Hossenfelder Physics

High temperature superconductors are badly understood theoretically, yet this understanding might allow us one day to create superconducting materials that save energy by avoiding resistive losses in long distance power lines. Imagine the potential! (Alternatively, read this.)

Presently the temperature at which these materials become superconducting is "high" only to the physicist who spends his days playing with liquid nitrogen: The transition temperature below which superconductivity sets in, also called the critical temperature, is in all known cases below -70°C. (The value depends on properties of the material as well as external fields.)

"Normal" superconductivity is described by the theory of Bardeen, Cooper and Schrieffer. At low temperatures, but in the not-superconducting phases, these metals are well described as Fermi liquids. But metals who display high temperature superconductivity are an entirely different story, and one that is largely unwritten.

One thing we know from experiment is that high temperature superconductors are "strange metals" whose electric resistance in the normal, non-superconducting, phase increases linearly with the temperature rather than with the square of the temperature. The latter is what one finds for a Fermi liquid with weakly coupled quasi-particles. Thus, plausibly the reason for our lacking theoretical understanding is that strange metals are strongly coupled system, which are notoriously hard to understand. "But darling," said the string theorist, "I can explain everything." And so he puts a black hole into an Anti-DeSitter (AdS) space and looks at the boundary.

The celebrated AdS/CFT correspondence makes it possible to deal with strongly coupled systems by mapping them to a weakly coupled gravitational system in a space-time with one more dimension. This is computationally more manageable, or at least one hopes so. So far, this correspondence, also called "duality", between the gravity in the AdS space and the strongly coupled theory on the boundary of this space (thus one dimension less) is an unproved conjecture put forward by Juan Maldacena. However, it has been extensively tested for a few cases and many people are confident that it captures a deep truth about nature (though they might disagree on the extent to which it holds).

For a high-temperature superconductor, one puts a planar black hole in the AdS space and decorates it with some U(1) vector fields and a scalar field, φ, and then goes on to calculate the free energy for different configurations of the scalar field. For temperatures above a critical value, the free energy is minimal if the scalar field vanishes identically. However, if the temperature drops below this critical value, configurations with a non-vanishing scalar field minimize the free energy, so the system must make a transition. In the figure below, you see the free energy, F, (with some normalization) as a function of the temperature (again with some normalization) for the case of φ = 0 (dotted line) and a case with non-vanishing φ (solid line). The latter solution doesn't exist for all values of the temperature. But note that when it exists, its free energy is lower than that of the φ=0 solution.

[Image credits: Hartnoll, Herzog and Horowitz]

For these different configurations one can then calculate thermodynamic quantities of interest, such as the electric conductivity (AC and DC) or heat conductivity, and… compare the results with actual measurements.

As you can tell already from my brief summary, this approach to understand strange metals is, presently, far too rough to give quantitative predictions. It can however describe qualitative behavior, such as the scaling of the resistance with temperature that is so puzzling. And that it does quite well!

A bunch of smart people have been studying the strange metal duals for a couple of years now, among others Subir Sachdev, Sean Hartnoll, Hong Liu (who wrote a recent article for Physics Today on the topic), Shamit Kachru, Gary Horowitz, and a group here at Nordita around Lárus Thorlacius.

Nordita Fellow Blaise Goutéraux is among the AdS/CFT correspondents of this group. He has taken on another challenge in this area, which is to describe the landscape of holographic quantum critical points, from which the strange metallic behavior at finite temperature is believed to originate. For this, Blaise works with more complicated geometries that exhibit different scaling behaviors from AdS.

What do we learn from this? The AdS/CFT correspondence is a tool, and if you've got a hammer quantum critical points start looking like nails. But the only reason we call the bulk theory gravitational is that we first encountered a theory of this type when we wanted to describe the gravitational interaction. Leaving aside this scientific history, in the end it's just a mathematical model to calculate observables that can be compared to experiment.

The big question is however whether this approach will ever be able to deliver quantitative predictions. For this, a connection would have to be made to the microscopic description of the material, a connection to the theories we already know. While this is not presently possible, one can hope that one day it will be. Then one could no longer think of the duality as merely useful computational tool with an educated guess for the geometry – the bulk theory would have to be a truly equivalent description for whatever is going on with the lattice of atoms on the boundary. But the cases for which the AdS/CFT correspondence has been well tested are very different from the ones that are being used here, and the connection to string theory, the original inspiration for the duality, has almost vanished. It wouldn't be the first time though that physicists' intuitions are ahead of formal proof.

Write a comment