SIO221a
Introduction to Data Analysis

FALL quarter

Instructor: Matthew Alford

Credits: 4

Time: M-W 11-12:20

Location: Spiess 330

Course Requirements:

Prior to class: please get set up with your analysis packages and Github for version control, as described in the syllabus.

Complete weekly problem sets. For most of the problem sets, you may work collaboratively, though the work that you submit must be your own. (Please follow the standards of scientific publication and identify your collaborators.) A midterm and final problem must be completed independently. (They will have about the same scope as the the other problem sets.)

The final problem set will be an independent project, which you will present during the final exam time slot. A draft write up will be due during the final week of classes, and the final write up of your project will be due no later than 11 am on the night of the final exam slot.

To gain from this class, students are expected to come to class, participate in class discussions, ask questions. There will be some assigned reading (available in electronic form), and students are expected to complete the reading.

The syllabus can be found here.

Schedule: see the UCSD academic calendar.

Topics:

Mean/standard deviation/pdfs— moments of pdfs - programming basics

Special PDFs - Gaussian, Chi-squared distribution - version control basics

Central Limit theorem, Error propagation

Correlation/covariance/projections onto modes

Fourier transform and spectra.

Spectral uncertainties

Windowing—degrees of freedom

Aliasing

Multidimensional spectra (e.g. Frequency/wavenumber spectra)

Coherence/cross spectra/transfer functions

Optional as time allows:  spectrogram rotary spectra, multitaper, maximum entropy, advanced filtering, filter design, Monte Carlo