Version 1.0.0

Reading and Folding ASM Data I

Reading and Folding ASM Data of Her X-1

All the examples presented here were created with the driver and example scripts found on the SITAR downloads page. To reproduce the results below, simply run the script as:

   isis> () = evalfile("fold_herx1.sl");

Note that you will have to have the Her X-1 data file from the All Sky Monitor web page, or which you can download from here.

The driver script is set-up to create data plots to the screen, as well as to echo the most essential commands that are being run, pausing inbetween. These defaults can be changed by editing the driver script. Below, we take a look at the output generated. (To see this repeated on the screen, type in ISIS or Sherpa > ()=evalfile("fold_driver");).

First, let's read in the Her X-1 data, only accepting 'ASM solutions' with chi^2 of better than 1.1, using the command

     ev = sitar_readasm("xa_herx1_d1.lc";maxchi2=1.1);

The results are placed in the ev structure (with fields .time, .rate, etc.). Plotting:

Her X-1 Unbinned

Clearly we'd like to bin that up a bit. Let's bin with 7 day wide bins, not worrying about binning up the ASM error bars (we'll later just use the sample error on the mean), and taking default start and stop times.

     rb = sitar_rebin_rate(ev.time,7,ev.rate);

The results are placed in the structure rb (with fields .bin_lo, .bin_hi,.rate, etc.). Plotting:

Her X-1 Rebinned

You can notice (or might have read) that there appears to be a distinct periodicity near 35 days. So, let's fold the data on this period. We choose to use a 35 day period, without any correction for first or second derivatives, and fold into 20 phase bins (the default), by typing:

     pf = sitar_pfold_rate(ev.time,ev.rate,35.);

The results are placed in the structure pf (with fields .bin_lo, .bin_hi,.mean, .sdm, etc.). Plotting, with error bars taken to be the standard deviation of the mean:

Her X-1 Folded

But what if you didn't know ahead of time that there was a 35 day period? One can search for periodicities using the epoch folding function. Here we search 500 periods (each folded into 20 phases), logarithmically spaced between 1 day and 100 days, by typing:

     ef = sitar_epfold_rate(ev.time,ev.rate,1.,100.; nphs=20,nsrch=500,loggrid);

The results are placed in the structure ef (with fields .prd, .lstat, etc.). Plotting:

Her X-1 Epoch Folded

One can clearly see significant peaks in the folding statistic at integer multiples of half of the 35 day period. Looking at the folded lightcurve above, with its two peaked structure (the "low" and "high" states), one can see why multiples of the sub-harmonic, and not just the main period, appear in the epoch fold.

This page was last updated Aug 22, 2011 by Michael Nowak. To comment on it or the material presented here, send email to mnowak@space.mit.edu.