topics.txt

* few body problem

  * energy conservation
  * sympletic integrators / Leap frog
  * close encounters / softening

  * look at Wisdom-Holman -- decomposes orbit into Keplerian part and interaction part.  See,
    e.g., https://academic.oup.com/mnras/article/452/1/376/1748797

* reaction networks

  * stiff ODEs

* stellar structure

   * ODEs
   * shooting

* self-gravity

  * multipole method

* hydro

  * subgrid modeling with Burgers equation

* fitting

  * orthogonal distance regression: https://docs.scipy.org/doc/scipy/reference/odr.html#module-scipy.odr


* bayesian

  see https://bayesiancomputationbook.com/welcome.html

  maybe create fake RV data from a system with 4 planets (with some inclination)
  and try to recover their orbital periods?

  here's a nice fit to a line example: https://arxiv.org/pdf/1008.4686.pdf

  Alex Ji: https://github.com/alexji/cda-samplers
           https://arxiv.org/abs/1802.06810

  Fitting a line: https://github.com/bmorris3/mcmc_interact

* machine learning

  instead of MNIST, try Cifar10

  sample dataset: 

     * galaxies: https://zenodo.org/record/4288837#.YYTamy2l1Z0

     * GZ2, GZ-DECaLs (Galaxy Zoo):

       * https://zenodo.org/record/3565489

       * https://zenodo.org/record/4573248

     See also: https://en.wikipedia.org/wiki/List_of_datasets_for_machine-learning_research#Astronomy

     Also the AstroML sample datasets: https://www.astroml.org/examples/datasets/index.html

  hello universe:

    https://archive.stsci.edu/hello-universe


* FFT

  for C++, consider using pocketFFT -- this is C++ header, and is used by NumPy
  https://gitlab.mpcdf.mpg.de/mtr/pocketfft/tree/cpp

  we can pass in the `.data()` of the underlying vector.

* Fitting an orbit:

  See this paper for grid search: https://ui.adsabs.harvard.edu/abs/1989AJ.....98.1014H/abstract
  and this: https://ui.adsabs.harvard.edu/abs/1999AJ....117.1023M/abstract

  nice example: https://docs.exoplanet.codes/en/v0.4.5/tutorials/astrometric/


* Gaussian processes:

  * review: https://arxiv.org/pdf/2209.08940.pdf

  * https://github.com/bmorris3/gp_interact