GModelFitViewer

The GModelFitViewer.jl package allows to visualize 1D GModelFit objects in a HTML page. Specifically, it allows to display a plot of the best fit model (along with all its component) and of the empirical data, as well as the logs reported in the Julia REPL when showing the objects.

The most relevant functions are:

• GModelFitViewer.serialize_json(): writes a JSON file containing the above mentioned contents;
• GModelFitViewer.serialize_html(): writes an HTML page containing the above mentioned contents;
• viewer(): writes an HTML page and displays it using the default browser;

Both functions share a syntax similar to the GModelFit.serialize() function. The accepted arguments are:

• a ModelSnapshot or a Vector{ModelSnapshot} object;
• a FitStats object;
• a Measures or a Vector{Measures} object;
• (optionally) a path and name for the output file via the filename= keyword. If the filename is not provided a standard one will be used and stored in the tempdir() directory.

Besides the above arguments, a number of optional keywords may be provided to customize the plot:

• title: plot title;
• xlabel: label for X axis;
• ylabel: label for Y axis;
• xrange: 2-element vector specifying the range for the X axis;
• yrange: 2-element vector specifying the range for the Y axis;
• xscale: numerical global scale factor for the X axis;
• yscale: numerical global scale factor for the Y axis;
• xunit: units for the X axis (as a string);
• yunit: units for the Y axis (as a string);
• rebin: an integer specifying the rebin factor along the X axis (X values are averaged, Y values are averaged using uncertainties as weights);
• keep: a String, a Regex, or a Vector{String}, indicating which components should be kept in the final file;
• skip: a String, a Regex, or a Vector{String}, indicating which components should be ignored when writing the final file;

The usage of rebin, keep and skip allows to produce files which are significantly smaller in size.

IMPORTANT NOTE: the keywords name may be abbreviated as long as the name in unambiguous. E.g., you may use xr in place of xrange, re in place of rebin, etc.

Create a GModelFit.jl model, generate a mock dataset and fit:

using GModelFit, GModelFitViewer
model = Model(:bkg => GModelFit.OffsetSlope(1, 1, 0.1),
              :l1 => GModelFit.Gaussian(1, 2, 0.2),
              :l2 => GModelFit.Gaussian(1, 3, 0.4),
              :main => SumReducer(:bkg, :l1, :l2))
dom = Domain(0:0.01:5)
data = GModelFit.mock(Measures, model, dom)
best, fitstats = fit(model, data)

Generate and display an HTML page vith:

viewer(best, fitstats, data);

You may customize the plot using the above mentioned keywords, e.g.

viewer(best, fitstats, data, 
       title="My title", xr=[0.5, 4.5], rebin=2, keep=r"l.")

To save the HTML page in myfile.html (without opening it in the web browser):

GModelFitViewer.serialize_html(best, fitstats, data, filename="myfile.html")