Full MCMC chain for 15,000 iterations for the example shown in Figure 2. Each of the small panels corresponds to one parameter. One sees that the “burn-in” region is confined to the first 1000 iterations. The estimated parameters are shown with the red line and are calculated from the last 60% of the chain. The gray lines show the ﹩1\sigma ﹩ standard deviations, and the dotted lines show the 95% confidence interval. Note the fitted flux value is ﹩(1.06\pm 0.03)\times {10}^{-16}﹩ erg s−1 cm−2, which is found from the sum of the pixel values (here ﹩\sim 8.5\ {10}^{-17}﹩ erg s−1 cm−2 Å−1) times the 1.25 Å per spectral pixel. The bottom panel shows the ﹩{\chi }^{2}﹩ evolution relative to the minimum, ﹩\mathrm{log}[{\chi }^{2}-{\chi }_{\mathrm{min}}^{2}]﹩. We use this nonstandard metric in order to show that the variations of the ﹩{\chi }^{2}﹩ around the minimum are 3–4 orders of magnitude smaller, reflecting a very flat hypersurface. Hence, a plot of ﹩{\chi }^{2}﹩ or of the likelihood would show a straight line.