Variability_report
A repository for a technical paper, that together with a Faint Pipeline Report accompany the data-driven research paper on SDSS Stripe 82 Summer 2013 reprocessing SDSS_S82_paper. We describe ways to distinguish variability from noise, ranging from time-domain (chi2, BIC, Bayesian parametrization) to frequency-domain (Lomb-Scargle periodogram).
Code used to make figures in this report is stored in the SDSS_S82_FP_research repo.
In particular :
| Research | Notebook |
|---|---|
| Chi2 variance simulation | Variability_chi2_test_AstroML |
| Weighted interquartile range, used in calculating the mean sigma | Variability_weighted_interquartile_range |
| What quantities are calculated for p(sigma) | Variability_descriptors_of_p_sigma |
| Chi2 tests, completeness curve | Variability_time_completeness_curve |
| Benchmarking AstroML 5.8 code: how number of bootstraps, number of points affect speed of execution | Variability_timeit_AstroML |
| Frequency : investigating gridding choice | Variability_frequency_grid |
| Frequency : choosing the best grid | Variability_frequency_grid_solution |
| Simulating the estimate of periodogram peak height from AstroML eq. 10.49 | Variability_frequency_AstroML_chap-10 |
| Comparing BIC to AIC in the periodogram context : AstroML Fig.10.15 | Variability_compare_BIC_vs_AIC_AstroML_Fig10-15 |