Examples: posterior parameter distributions and spectra (inlaid at upper right) for an AGN (ID = 34) fitted with an absorbed power-law (left) and for a star (ID = 50) fitted with an APEC model (right). The posterior distribution of each individual parameter and each pair of parameters are plotted using the “corner” package (Foreman-Mackey 2016), with the median and 68% percentile interval around the median printed on top of each individual distribution. The spectral shape parameters for the absorbed power-law model include the slope (Γ) of the power-law and the logN_H of the intrinsic absorber (in cm^–2). For the APEC model, the spectral shape parameters include the Galactic absorption column density logN_H, the temperature log KT (in ke V), and the abundance. All the normalization parameters (e.g., logPowNorm, logApecNorm) are measured in logarithm space. The background normalization parameter (logBkgNorm) is set free in the fitting and always equals one. In the spectral plot, the data (dark blue points) are rebinned just for representation to reach an S/N of 3 but allowing at most 6 adjacent bins (each bin includes 4 channels) to be grouped. The blue solid, dashed, and dot-dashed lines display the best-fit model folded with the instrument responses, the folded best-fit source model, and the best-fit background model rescaled to the source extraction region. The shaded region indicates the 99% percentile confidence interval of the model around the median. The lower panel compares the data with the best-fit model in terms of (data-model)/error, where the error is calculated as the square root of the model predicted number of counts.