New features
- Added new models with a velocity broadening parameter to the
cph, equil, gadem, gnei, nei, npshock, pshock, sedov, snapec, and wdem
families. For instance bcph is the cph model but with a velocity
broadening parameter and bvcph is the vcph model with a velocity
broadening parameter.
- Added new models vvgadem and bvvgadem with all parameters for
all abundances.
- Added new models with rationalized names although the old names
are still available. So (b)(v)cempow can be used instead of cemekl
and cevmkl; (b)(v)cheb6 instead of c6mekl and c6vmekl;
(b)(v)expcheb6 instead of c6pmekl and c6pvmkl; (b)(v)coolflow
instead of mkcflow and vmcflow.
- All the collisional ionization equilibrium models which use
atomdb data can now also use data from spex v3.07.03. The spex data
can be used in two ways: 1) use xset APECUSESPEX yes in which case
everything applies as usual except that the SPEX tables are use; 2)
for those models with a switch parameter spex can be chosen by
setting the value to 3. In this second case the equivalents of the
apec control variables are available, eg SPEXROOT, SPEXNOLINES, etc.
- Added new (b)(v)(v)cie models to allow easy switching between
atomic physics sources for a single temperature plasma. These are
equivalent to the apec models with a switch parameter added.
- Added code to calculate electron-electron bremsstrahlung using
formulae from Nozawa et al. 2009. This is available either as a
separate eebremss model or included in the apec-based models using
xset APECEEBREMSS yes.
- Set the default redshift for all cooling flow models to 0.1
since 0.0 does not work.
- Set the switch parameter to 2 (ie atomdb/apec) for all models
where this parameter determines which atomic physics to use.
- Improved output from the identify command.
- Added to Numeric.h the mass of the e- in keV, the Thomson
x-section, and the fine structure constant.
- Added ability to specify user-defined values for the contour
levels of 1-D contour plots.
Bug fixes
The following patches for 12.13.1 have been included:
- 12.13.1a. Table models with both redshift and escale set are not
handled correctly.
- 12.13.1b. When an MCMC chain file contains response parameters, Xspec
is falsely reporting mismatches between the chain parameters and the
current variable fit parameters. Therefore it doesn't allow the
chain to be used for uncertainty estimates in place of fit
covariance values.
- 12.13.1c. In PyXspec's AllModels.addPyMod method, usage of a
deprecated Python function is preventing execution with the latest
Python v3.11.
- 12.13.1d. There is an error in the calculation of the pgstat
statistic.
- 12.13.1e. The vvapec, bvvapec, vvtapec, and bvvtapec models all deal
with the H abundance parameter incorrectly. If it is not zero, it is
reset to one. So, e.g. setting the H abundance to 0.5 in vvapec
actually produces a spectrum with the H abundances as 1.0. If the H
abundance is set to 1.0 (the default and usual setting) the results
are correct.
In addition the following bugs have been fixed:
- Fixed error in ISOTROPIC bayes prior - it should be sin, not cos.
- Fix to the case of 2-pane plots which contain a plot group
with zero channels.