Attention!

Probably need to put a suited-up picture here.

I am an associate project scientist at Berkeley Center for Cosmological Physics. I develop and maintain the software stack that supports the research at BCCP, including my own research.

I develop a lot of software packages, and contributes to many more. Most of these can be found at my github page. In earlier days I was involved with GNOME2 with a side-line project gnome-globalmenu; I do not think many people still use it; as the features have been partially included in GNOME3 and Ubuntu Unity.

My professional work spans three fields:

  • cosmology,
  • massively parallel super-computing applications, and
  • methods and software tools for data analysis;

Please refer to my CV; or a list of publications.

Here is a list of research topics and projects that I have contributed to:

  • Cosmology
    • Large Scale Structure (LSS) and perturbation theory; [5] [7] [18] [8]
    • Formation of galaxies, galaxy clusters, star-formation models and super-massive blackholes. [2] [1] [3] [4]
    • Weak lensing and intrinsic alignments. [6]
    • Reduction of imaging data from imaging surveys. [22]
  • Super-computing Applications
    • Solver for cosmological hydrodynamics; [9]
    • Parallel Fast Fourier Transformation; [10]
    • Massively parallel sorting; [11]
    • Efficient launching of Python applications on super-computers; [12]
  • Software and methods for data analysis
    • Automated differentiation [20]
    • Optimization beyond first order [21]
    • Parallel data analysis of N-body simulations [13]
    • In-node Map-Reduce via Multi-processing [17]
    • Hierarchical Data Format for large data sets [19]
    • KD-tree, pair-wise counting and Friends-of-Friends [14] [15]
    • Visualization of smoothed particle field [16]

I was part of several largest cosmological simulations. These simulations required some of the largest super-computers today(or then), running at their full capability for days; that's hundreds of millions of CPU-hours. We do make important predictions about blackholes and galaxies with these simulations.

[1] High-redshift supermassive black holes: accretion through cold flows. http://adsabs.harvard.edu/abs/2014MNRAS.440.1865F
[2] Cold Flows and the First Quasars. http://adsabs.harvard.edu/abs/2012ApJ...745L..29D
[3] The MassiveBlack-II simulation: the evolution of haloes and galaxies to z ˜ 0 . http://adsabs.harvard.edu/abs/2015MNRAS.450.1349K
[4] The Formation of Milky Way–mass Disk Galaxies in the First 500 Million Years of a Cold Dark Matter Universe. http://adsabs.harvard.edu/abs/2015ApJ...808L..17F
[5] Mock Quasar-Lyman-α forest data-sets for the SDSS-III Baryon Oscillation Spectroscopic Survey. http://adsabs.harvard.edu/abs/2015JCAP...05..060B
[6] Intrinsic alignments of galaxies in the MassiveBlack-II simulation: analysis of two-point statistics. http://adsabs.harvard.edu/abs/2015MNRAS.448.3522T
[7] Eulerian BAO Reconstructions and N-Point Statistics. http://adsabs.harvard.edu/abs/2015arXiv150806972S
[8] Perturbation theory, effective field theory, and oscillations in the power spectrum http://adsabs.harvard.edu/abs/2015arXiv150902120V
[9] http://bluetides-project.org/coderelease/
[10] Python binding of a library for massively Parallel Fast Fourier Transform (PFFT). http://github.com/rainwoodman/pfft-python
[11] A library for parallel histogram sort with a Python binding. https://github.com/rainwoodman/MP-sort
[12] Boosting python start-up speed with MPI. https://github.com/rainwoodman/python-mpi-bcast
[13] Parallel Data Analysis for N-body simulations. http://github.com/bccp/nbodykit
[14] Spatial data with a KD-tree. http://github.com/rainwoodman/kdcount
[15] Scipy cKDTree rewrite (pending, https://github.com/sturlamolden/scipy/pull/1)
[16] http://github.com/rainwoodman/gaepsi
[17] http://github.com/rainwoodman/sharedmem
[18] A pure particle mesh solver. http://github.com/rainwoodman/fastpm
[19] Hierarchy Data Files. http://github.com/rainwoodman/bigfile
[20] A simple autodiff and a small library of operators. http://github.com/rainwoodman/vmad
[21] (ABstract OPTimizer) - optimization of generic numerical models. http://github.com/bccp/abopt
[22] http://desi.lbl.gov