Smoothed data plotting, and comparison to rebinning.

dph Sun Oct 28 2001

The file, splot_data.sl contains the functions, splot_data_counts(hist, sigma) and splot_data_flux(hist, sigma) which apply a Gaussian kernel of dispersion (in Angstroms), sigma to the data before plotting, using the function, gsmooth.sl. The figure shows a comparison of data smoothed by the bin size (0.005A for MEG and 0.0025A for HEG) in the top panel. The dark blue histogram is the MEG spectrum, and the green is the HEG. The overplotted red histogram is the smoothed result.

The bottom panel shows the same data before and after rebinning by 2 pixels, and then multiplying by 0.5 to scale to the un-rebinned counts.

Smoothed data should NEVER be used for quantitative analysis, since the statistics are no longer easily interpreted. However, it might be considered to give a more aesthetically pleasing representation of a smooth underlying model, and could be used for presentation, or inspection.
The plots shown were made with Smooth_rebin_comparison.sl.