Autumn 2015

For the Autumn 2013 course web page, go here.

Assignments are from Numerical Recipes (Press et al.) or Statistics, Data Mining, and Machine Learning Astronomy (Ivezic et al.).

I have written a reader's guide to chapters 3-5 of Ivezic. It will be useful to read the guide before you read the corresponding chapter, and maybe to read it again afterward, to get the overall context of what is covered in these three long and wide-ranging chapters.

- Reading 1: Read NR Chapter 4, mainly 4.0-4.4, on numerical integration. Also read the Introduction and Appendix A of Statistics ...
- Subsequent assignments were given in the notes or homework assignments. Sorry I didn't keep up the list here.

- Numerical Notes 1: Numerical Integration
- Statistical Notes 1: Some High-Level Background
- Numerical Notes 2: Ordinary Differential Equations
- Numerical Notes 3: Root-Finding and Minimization
- Numerical Notes 4: FFTs and Their Applications
- Statistical Notes 2: Bayesian Parameter Estimation
- Statistical Notes 3: Covariances, $\chi^2$, MCMC, and Fisher Matrix
- Statistical Notes 4: Observables and Parameters
- Statistical Notes 5: Hypothesis Testing
- Statistical Notes 6: Estimating Error Bars From Data

- Problem Set 1: Numerical Integration,
due Tuesday, 9/15.
- Starting code: integrate.py and integrate_sub_starter.py.
- Simple wrapper for running Python integrators, si.py.
- sm.converge, SM script for convergence plot.

- Problem Set 2: Numerical Integration of Orbits,
due Tuesday, 9/29.
- Starting code: orbit_euler.py.
- Plotting scripts: sm.orbits.kep, sm.energy.kep, sm.orbits.ho, sm.energy.ho, sm.orbits.moon.
- Shell scripts for running a sequence of Kepler and H.O. runs: sh.kepruns, sh.horuns.

- Problem Set 3: Root Finding and Minimization,
- Useful code for Part 4: cosmodist.py and cosmodist_subs.py.

- Problem Set 4: Multi-variate Gaussians and Simple MCMC
- Example code for multi-variate Gaussian: gauss2d_example.py and

- Problem Set 5: Forecasting
- Codes, data files, and plotting routines for PS 5 (described in Problem Set)
- linedata.py
- line_mcmc.py
- linepluscov.py
- line.n20.s12.dat
- line.n20.s17.dat
- line.n20.s0.dat
- line.n6.s0.dat
- sm.linemc
- sm.linepluscov

- Problem Set 6: Fun with Fourier Transforms
- Codes and data files for PS 6
- fftsimple.py
- fftnoise.py
- gaussft.py
- data1.out for Part 3.
- data2.out for Part 3.

- Problem Set 7: Systematics and Nuisance Parameters
- Data file for PS 7: h0.data

- Some resources for the SM plotting package:
- A meta-introduction to SM.
- A tutorial on SM
- .sm environment file. This should go in your home directory, called .sm, with all /home/cosmology/dhw occurrences changed to your home directory.
- DHW's macros. Given the dotSM file, this should go in a directory Smacro in your home directory, with filename "default."
- DHW's graphcap definitions. This should go in the Smacro directory with filename "privcap."
- A simple example of an sm plotting script. You need to install the macros above for it to work. If you make this script executable, then "sm.example 5" will put a plot on your screen for 5 seconds (you can ^Z to keep it there longer), and "sm.example 0" will put the plot in the file "sm_example.ps".

- A quick reference guide for Python plots, from Cassi Lochhaas.
- Intro to Python slides from Demitri Muna.
- Tom Loredo's Bayesian Reprints. The first two articles on this page, "From Laplace to SN 1987a ..." and "The Promise of Bayesian Inference ...", are excellent introductions to Bayesian statistical methods. There is substantial overlap between them; the first is more "philosophical" in orientation, the second more "practical."

Go to David Weinberg's Home Page

Updated: 2015 December 8[dhw]