Portfolio

I have worked on various programming projects over the years, this page is an overview over the projects. The portfolio showcases the most important ones, further below is a simple list of the other projects.

Lüscher Analysis

For the $I = 1$, $I_3 = 1$ scattering project I created most of the analysis code using R and our hadron library. The analysis is written in a general way that we can re-use most of the infrastructure for later projects, hence the general name of the code. It starts with the projected correlation functions and does all these steps:

  1. Outlier detection
  2. Estimation of autocorrelation time
  3. Error estimation with bootstrap and/or jackknife
  4. Generalized eigenvalue problem (GEVP) solving
  5. Effective mass analysis
  6. Fitting of two-point and four-point correlation functions
  7. Conversion of energies into phase shifts via the Lüscher method
  8. Fitting of Breit-Wigner and IAM phases
  9. Extrapolation in lattice volume, lattice spacing and pion mass

Each of these steps feeds into extensive R-markdown documents that render to documents with many plots and tables.

Parameter Value Framework

In data analysis I have often faced the need to try multiple variants of the same steps. Inserting variants here and there quickly lead to unmanageable code. The parameter value framework paramvalf provides an effective mean to manage arbitrary many variants of the same analysis. A table with parameters is used to distinguish all the variants and associated values. Different parameter-value-objects are combined using SQL join logic to automatically cover all parameter combinations that are sensible and provide the user with the matching values.

The current implementation is done with R and makes heavy use of the dplyr library. The same ideas could be implemented in Python as well.

hadron

The hadron library is our Lattice QCD analysis toolbox that has functionality for IO, error estimation, GEVP solving, fitting, plotting and other tasks. It is written in R. Since 2018 all members of the group use this for their analysis projects and most people contribute to the library.

My contributions are some new functions but mostly refactoring that was necessary in order to cleanly extend the library. Also I introduced unit tests with testthat and continuous testing on Travis CI. Enforcing that all changes to the master branch go through a reviewed pull request has improved our joint work on the library.


sLapH Projection NG

There was a sLapH projection code that was able to do pion–pion scattering but only with two particles, only in $I = 1$, $I_3 = 1$ and only in P-wave. For my project with three pions I needed something that could project other operators down to the lattice irreducible representations of the octahedral group and created sLapH-projection-NG.

The group theory was already figured out, so I “just” needed to implement this in its general form. For this I have chosen the Wolfram Language and Mathematica. Using this code we can now project operators with arbitrary many particles to all the lattice irreps. During this project I learned a lot about the Wolfram Language and were able to use my functional programming knowledge from Haskell.

There also is a second part which does the numerical projections using the prescriptions computed analytically. This part is written in R and highly parallelized such that each irrep and configuration can be run independently of each other.

sLapH Contractions

One step in the lattice QCD measurement process are the contractions. Large tensorial quantities are reduced down to correlation functions. Tensor contractions themselves are not a hard problem, the logistics around them are. I have contributed to the existing C++11 project sLapH-contractions with a major refactoring. The first step has been the introduction of an integration test such that between each chance we were certain that the results have not changed and have them run on Travis CI. This allowed complete overhaul of control flow without an endless bug search later on.

After the refactoring, there is less code, more features; it uses less memory and further additions are now easier than before.

The project uses OpenMP for multithreading, Eigen for the actual matrix multiplications, CMake to build and HDF5 to store the results.


QPhiX

As part of my master thesis, I have worked on QPhiX, a solver for sparse systems of equations as they appear in lattice quantum chromodynamics (Lattice QCD). It is written in C++ and uses MPI for inter-node communication and OpenMP for threading. The code is tailored for the Intel Xeon Phi platform which runs 64-Bit x86 code but has a large number of threads. The various SIMD length of the Xeon and Xeon Phi architectures requires a data structure that is flexible enough.

A code generator generates intrinsics code for the various targets (none, SSE, AVX, AVX2, AVX512) which by construction unrolls all loops and can interleave software prefetches (needed for Knights Corner). The sparse matrix multiplication is memory bandwidth bound, QPhiX achieves some 75% of the theoretical performance on a single node.

My contributions to the project:

  • Extension to support two-flavor operators, which is effectively the addition of another dimension to all the array structures. A lot of C++ templates have been used to offer both one-flavor and two-flavor operators without changing existing interfaces and prevent performance regressions.

  • Introduction of continuous integration on Travis CI. All code is build for general architecture, AVX, AVX2 and AVX512 for each commit in git. Unit tests are run with MPI for AVX on Travis CI.

  • The build system was ported from GNU Autotools to CMake. The code generator was included in the main repository and code is now generated during the compilation using Jinja templates.

SU(2) Hybrid Monte Carlo

As part of my master thesis, I have written a Hybrid Monte Carlo (HMC) simulation for SU(2) Yang-Mills theory (just gluons, no fermions with two colors). It was an exercise to get to know the HMC algorithm by Duane et al and was compared to data by Creutz. It is available at su2-hmc.

It uses Eigen for the matrices, Boost for its INI parser, gtest for unit tests, and OpenMP for parallelization of loops. Also it uses CMake to build.


Fast Multipole Method

Within the scope of the Guest Student Program (GSP) at the Jülich Supercomputing Center (JSC), I got to work with the Fast Multipole Method (FMM) group for a little over two months. The FMM is a scalable algorithm to compute the long ranged forces between particles with computational complexity linear in the number of particles.

The previous Fortran implementation had been rewritten in C++11 to use costless abstraction features like templates and make specialization for different architectures manageable. During my stay I parallelized one step of the whole algorithm using a communication optimal algorithm by Driscoll et al using MPI. Scaling tests have been performed on the BlueGene/Q installation JUQUEEN.

MEGraMa Data Analysis

The DLR Institute of Materials Physics in Space conducts a series of experiments of granular matter in weightlessness. Particles, in my case about 100 steel spheres of a few millimeter diameter, are excited using magnetic fields and then let to cool down; all while the experiment drops in Bremen’s drop tower. Three high speed cameras observe the weightless particles from perpendicular directions. In less than 10 seconds, five gigabyte of footage has been recorded.

My task has been to extract the 3D trajectories of the particles. In a four month project I have used the OpenCV library to extract 2D information from the images and a 3D model for the cameras (focal length, position, orientation). Correlating the data obtained from the three cameras allowed me to reconstruct the 3D path that the particles took. In this process several [Boost]http://www.boost.org/ classes have been used as well as linear algebra via Eigen. Using those 3D trajectories, the group has been able to compute interesting physical quantities.

The process of extracting data needs a lot of parameters. Thresholds have to be tuned for optimal image processing, path reconstruction, and 3D correlation of data. In order to make this task productive, I have added a Qt 4 GUI that presents the parameters to the user. Leveraging the performance of C++ with OpenMP allows to see the results in real time on a standard desktop computer.


thinkpad-scripts

My ThinkPad X220 Tablet has a rotatable screen which includes a Wacom digitizer pen. On desktop Linux, neither the special bezel buttons nor the correct digital rotation when physically turning the screen works out of the box. In order to conveniently use the laptop, I have started with a small XRandR script in Bash and expanded it to correctly map everything when external screens are present, disabling the TrackPoint, and adding support for the docking station.

A couple years ago I rewrote the collection of Bash scripts in Python. This allowed contributors and me to supply configuration file support, more features for edge cases, and also write unit test. The thinkpad-scripts project is mature now and has attracted a few forks and pull requests.

Unfortunately it only works with X.Org and I have not looked into making it work with Wayland.

jscribble

At first I did not find a note taking program that I could use with my netbook and Wacom digitizer tablet. Taking handwritten notes in a normal photo editing program is cumbersome as one has to create new images for pages manually. My need was to create a new page with one button and quickly switch between the pages.

Using Java and combining a MouseMotionListener with the JPanel.paintComponent led to the very basic drawing program jscribbe that allowed me to take as many notes as I wanted. Today I use Xournal which is a much more sophisticated program for taking handwritten notes.


The whole list

The following is a terse list of the projects that I have worked on over the years. The links usually point to the GitHub repository where you can find more information about the projects as well as the source code.

Admin

Development

Finance

Games

Some utilities for Urban Terror:

Graphics

Hardware

LaTeX

Science

Lattice QCD:

Fractals:

Group Theory:

Data Analysis:

  • fastplot: Quickly generate X-Y plots from TSV data
  • format-value-error: Pretty printer for value and error, implemented in Octave, Python, and R
  • paramvalf: Parameter value framework
  • test-fit-error: Test error scaling of different curve fit implementations

Sound & Video

Utilities

Websites and Webapps