PriorΒΆ
gsf calculates posterior probability through an MCMC process, whose concept is based on the Bayes theorem.
For each parameter set in the minimizer, prior is set (posterior_flexible.py). In the current version, prior for most of the parameters (amplitude, metallicity, dust attenuation, redshift) are set flat, so that;
\[\begin{split}prior(\theta) &= 1\ \mathrm{if}\ \theta_\mathrm{min} < \theta < \theta_\mathrm{max} \\
&= 0\ \mathrm{else}.\end{split}\]
Below is the list of parameter ranges set in gsf.
Parameter ranges:
| Parameter | Range | Description |
|---|---|---|
| A | [AMIN:AMAX] | Amplitude, in logarithmic space. AMIN and AMAX are set to -3 and 3 in default. |
| Z | [ZMIN:ZMAX] | Metallicity, in logarithmic space. |
| AV | [AVMIN:AVMAX] | Dust attenuation in V-band, in linear space. AVMIN and AVMAX are set to 0 and 4 in default. |
| z | [ZGAL-EZL:ZGAL+EZH] | Redshift, in linear space. EZL and EZH are set to 0.3. Can be specified in the configuration file. |
| AGE (tau-model) | [AGEMIN:AGEMAX] | Age of tau model, in logarithmic space. Range can be specified by AGEMIN and AGEMAX. |
| TAU (tau-model) | [TAUMIN:TAUMAX] | Age of tau model, in logarithmic space. Range can be specified by TAUMIN and TAUMAX. |