Plotting ToolsΒΆ

There are several scripts for making diagnostic plots and visualizations. Some are described here.

  • Make a ds9 region file from an input fits catalog (use region_file_fits) or a list of artificial stars (use region_file_txt). Can also choose a column+value as a cut to set two different region colors.

  • Make a histogram of the AST fluxes for each filter. Optionally, also include a histogram of the SED grid for comparison.

  • Make a histogram of the best chi2 values (chi2=1 and the median chi2 are marked). Note that there is no plot of reduced chi2, because it is mathematically difficult to define the number of degrees of freedom. Inputs are the BEAST stats file and optionally the number of bins to use for the histogram.

  • Make a color-magnitude diagram of the observations. Inputs are the photometry file (which can be a simulation) and the three filters.

  • Similar to above, but color-coding the data points using one of the parameters from the BEAST fitting. Takes three additional inputs: a BEAST stats file, the parameter to use, and whether to apply color after taking the log10 of the parameter.

  • Make a triangle plot with completeness averaged into 2D plots for each pair of parameters, and 1D plots along the diagonal for each individual parameter.

  • For a given star, makes a multi-panel plot that shows the PDFs and best fits of each parameter, as well as an SED (similar to Figure 14 in Gordon+16).

  • For a given star, makes a triangle plot with all of the 2D PDFs. Diagonals contain the 1D PDFs.

  • Make histograms of the magnitudes for each band in the photometry catalog.

  • Plot the bias and uncertainty as a function of flux (similar to Figure 12 in Gordon+16). Multiple noise models can be overplotted, as long as they correspond to the same SED model grid.

  • Reproduce the Figures 16-18 in Gordon et al. 2016. Make 2D histogram of 50th percentile values of each parameter against its uncertainty (=0.5x(percentile_84th-percentile_16th)) on the left columns. Make H-R Hess diagram colored coded by the uncertainty of a given parameter on the right column.

  • Make a 2D histogram to compare simulated and recovered model parameters (similar to Figure 13 in Gordon+16). If given multiple sets of files, can do additional panels to compare across noise models.

  • Make a triangle/corner plot of all the parameters (p50) against each other. Diagonals contain histograms of each parameter.