Since the release of version 0.12.0, we have identified a few bugs due to the important architectural changes during the previous development cycle. Version 0.12.1 fixes the photometric redshift mode, fixes the estimation of the physical properties when some invalid models are present, and fixes a crash when plotting the PDF.
It has been a very rich and exciting year for Cigale and a new release was long overdue. This wait was really worthwhile and this new version 0.12.0 comes with numerous improvements of all sorts. Among the many notable changes:
- Dust templates generated with THEMIS (Jones et al. 2017) have been contributed by the DustPedia team (Davies et al. 2017). Special acknowledgement to Angelos Nersesian and Frédéric Galliano for creating the dust templates and writing the corresponding module.
- High resolution BC03 models have been added. They can be activated when building the database by adding –bc03res=hr.
- It is now possible to compute the models by blocks when there is not enough memory to compute all of them in one go. There is only a very marginal slowdown.
- The rapid evolutionary phases of stellar populations lasting less than 1 Myr should now be better captured.
- It is now possible to compute additional properties such as the bolometric stellar luminosity.
- Small speed optimisations should make models ~1% faster to compute.
In addition there have been a some under-the-hood modifications in preparation for the new great things coming to Cigale in the next months. Stay tuned!
The development of this new version of Cigale was partially supported through grant FONDECYT REGULAR 1170618.
- Nebular models have been improved, refined, and extended towards lower radiation field intensities.
- There is a new module to compute rest-frame quantities such as the UV β slope, IRX, Dn4000, emission lines equivalent widths, and colours & luminosities in any set of bands.
- The handling of filters has seen various improvements and a new set of filters has been added. Run “pcigale-filters list” for a complete list.
- Several optimisations have led up to ~30% in speedup for the computation of the models.
- Fix for a crash with Numpy 1.12 and another one when using upper limits.
- The minimum python version has been bumped to 3.5, however for better performance we recommend using Python 3.6.
Finally as usual we strongly suggest that you regenerate your pcigale.ini file. Let us know if you have any question or comment.
- The output files are now always generated in the form of ASCII and FITS tables
- A crash has been resolved when using the m2005 module
- The models are now always computed in the same order, which is particularly useful for comparing different runs with the savefluxes module for instance.
- Various speed optimisations making the code 10-15% faster than before.
Note that as usual, we strongly recommend that you regenerate your pcigale.ini file.
- It is now possible to pass the parameters through a flat file (one line for each model and one column for each parameter).
- The input parameters now undergo a sanity check to ensure that they do not create problems.
- For convenience the output files have been merged into a single file. The parameters are prefixed with “bayes” when estimated from the PDF and with “best” if they come from the best fit. This file is provided both in FITS and ASCII formats.
- For efficiency, the output spectra and SFH are now saved in FITS rather than in VO-table.
- The behaviour of dustatt_calzleit has been modified to ensure that the actual E(B-V) corresponds to that requested in input. The results are unchanged but it means you may need to sample the E(B-V) to higher values now.
- Various fixes to improve the accuracy of the results, in particular when z>0.
Note that pcigale.ini generated with older versions of cigale do not work anymore. They will need to be regenerated with this version.