Running statistical checks with `pygwb_stats`

The pygwb.statistical_checks module provides a way to visualize the results of an analysis run. Through a series of plots, it offers the possibility to check the results for statistical consistency. The module itself is detailed in the module documentation. Here, we provide some information on how to run pygwb_stats, a script that calls the statistical checks module and generates all the relevant plots.

1. Script parameters

To obtain information about the parameters of the pygwb_stats script, one can execute the following command:

pygwb_stats --help

This displays the optional arguments that can be passed to the script:

-c COMBINE_FILE_PATH, --combine_file_path COMBINE_FILE_PATH
                      combined file containing spectra
-dsc DSC_FILE_PATH, --dsc_file_path DSC_FILE_PATH
                      delta sigma cut file containing sigmas and more
-pd PLOT_DIR, --plot_dir PLOT_DIR
                      Directory where plots should be saved
-pf PARAM_FILE, --param_file PARAM_FILE
                      Parameter file used during analysis
-fs FONT_SIZE, --font_size FONT_SIZE
                      Primary label font size
-fcoh COHERENCE_FILE_PATH, --coherence_file_path COHERENCE_FILE_PATH
                      Path to coherence file. If passed, automatically triggers the plot coherences option.
-t TAG, --tag TAG     Tag to use when saving files
-co CONVENTION, --convention CONVENTION
                      Overall convention to use in plots

A brief explanation of the parameters is given in the code snippet above, and we provide some additional information in the section below when calling the pygwb_stats script.

2. Running the script

After running the pygwb_combine script (as explained here), the script will have produced a file that looks similar to

point_estimate_sigma_spectra_alpha_0.0_fref_25_t0-tf.npz

This file contains the combined spectra, where the notation indicates it was run with a spectral index of 0, reference frequency of 25 Hz, and t0 and tf would be actual numbers corresponding to the start and end time of the analysis, respectively. This file should be passed to the pygwb_stats through the -c argument.

If the pygwb analysis was run with the delta-sigma cut turned on, a file delta_sigma_cut_t0-tf should be present in the output directory as well. This can be passed through the -dsc argument.

The directory where the plots should be saved is passed through -pd, whereas the parameter file that was used during the pygwb analysis, i.e., the one passed to pygwb_pipe, should be passed through -pf. The label font size can be changed, by using the -fs option. Finally, the coherence file, if computed during the pygwb_pipe run, can also be passed via -fcoh, which will create a series of additional plots related to the coherence.

The pygwb_stats script can be run using, for example, the following line of code:

pygwb_stats -c {my_combine_file_path} -dsc {my_dsc_file_path} -pd {my_plotting_directory} -pf {my_param_file_path}

The script produces some output, which is saved in the output directory specified through the -pd argument.

3. Output of the script

The output of the pygwb_stats script contains a series of plots, atuomatically saved to png format, in the provided plotting directory -pd. The plots follow the naming convention {ifo_1}{ifo_2}-{t0}-{tf}-{content_of_plot}.png. Each of these plots illustrates some quantities that provide insights on the statistical quality and behavior of various quantities of interest of the analysis. We refrain from giving a plot by plot discussion of the various figures here, and refer the user to the dedicated demo on this topic here. In addition, several quantities of interest are saved to file for further follow-up (see here).

Running statistical checks with pygwb_stats

1. Script parameters

2. Running the script

3. Output of the script

Running statistical checks with `pygwb_stats`