Folding RR Lyrae data

In the previous class I introduced the RR Lyrae phenomenon. Briefly, RR Lyrae are pulsating stars of spectral class A. These stars have played a major role in the development of modern astronomy. The absolute magnitude of RR Lyrae stars is about 0.75 (about 40 times brighter than the Sun). What makes them interesting is that there exists a relation between the pulsation period and the absolute magnitude (see presentation in previous class). Thus merely measuring the pulsation period of an RR Lyrae (when combined with the apparent magnitude) immediately informs you of the distance to the star. RR Lyrae stars allow us to measure the distances to Globular clusters (and thence set the scale of our Galaxy).

The periods of RR Lyrae range from 0.2 to 1 day. A collection of "time series" for a number of RR Lyrae stars (the data is from the Palomar Transient Factory or PTF) can be at RRLyrae. For each RR Lyrae there are two files, the .txt file (data) and a .jpg (pulse profile). The data files looks like 

54907.843191 16.026 0.022
54907.941112 16.084 0.028
54961.730973 15.881 0.009
54961.788813 15.957 0.009
54961.839863 16.004 0.008
54961.904593 16.049 0.009
The first column is the MJD and is followed by R-band magnitude and error in magnitude. The periods of the objects can be found in periods.txt. My MATLAB program can be found here and the output can be found at here.

Home/Class-work:

  1. The first task in analyzing any time series data is to inspect the data. So plot the data as a function of sample index or MJD (or both). Next, histogram the amplitude and the associated error bars.
  2. Next read in the txt files the txt files and fold the data at the period and plot the folded curve. Compare your folded light curve with the corresponding .jpg file.
  3. Real data, almost always, has glitches or anomalies (e.g. cloudy night). Plot the data and eliminate "outliers". There is a science & art to it.
  4. Fold the clened up data.
  5. Optional: Repeat this exercise but use "optimal addition" and compute the error bars on the folded light curve.