Time series analysis in Python for physics and astronomy 0800-SZERCZAS
1. The basic structure of time series, including constant level, trend, periodic component, noise, and signal-to-noise ratio, as well as their standardization and normalization (mean and standard deviation).
2. Autocorrelation of time series.
3. Cross-correlation of two time series. Pearson correlation coefficient. Spearman rank correlation coefficient.
4. Smoothing and filtering: moving average
5. Smoothing and filtering: Savitzky–Golay filter
6. Fourier (harmonic) analysis — discrete Fourier transform.
7. Smoothing and filtering: low-pass filter, high-pass filter, and Wiener filter
8. Periodogram.
9. Window function and aliasing.
10. Periodogram analysis of unevenly sampled data
11. Types of periodograms used in data analysis.
12. Time series prediction: ARMA and ARIMA models
13. Gaussian processes.
14. Regression using Gaussian processes.
15. Modeling with Gaussian processes.
Total student workload
Learning outcomes - knowledge
Learning outcomes - skills
Learning outcomes - social competencies
Teaching methods
Expository teaching methods
- participatory lecture
- programmed material
Exploratory teaching methods
- classic problem-solving
- laboratory
- brainstorming
- project work
Type of course
Prerequisites
Course coordinators
Assessment criteria
The final grade will be a weighted average of the grades earned for (1) active participation in class (40%) and the final Jupyter notebook project (60%).
Grading criteria:
- 0-49% grade 2,
- 50-59% grade 3,
- 60-69% grade 3+,
- 70-79% grade 4,
- 80-89% grade 4+,
- 90-100% grade 5.
Bibliography
Robert K. Otnes, Loren D. Enochson, "Digital time series analysis"
Additional information
Additional information (registration calendar, class conductors, localization and schedules of classes), might be available in the USOSweb system: