Below you will find a list of known ctools issues. You may also check the list of known GammaLib issues.
Model fitting issues
Certificate problem when cloning from Git
When cloning ctools from git you may encounter an SSL certificate
problem. This is related to the usage of the https protocol and can
be resolved by setting the
GIT_SSL_NO_VERIFY environment variable
ctools does not compile against conda Python
Trying to compile ctools against conda Python may fail due to incompatibility issues. If you’d like to compile ctools against conda Python, make sure that gcc, swig, and cfitsio are all installed via anaconda
$ conda install gcc swig $ conda install -c conda-forge cfitsio
Python module does not work
ctools include a Python module that is built from so called wrapper
files that are autogenerated using the swig
tool. These wrapper files are shipped with a ctools release, but if
you use the code from git you need swig
to generate the wrapper files during the build step. In any case,
to compile the Python module ctools need the
Python.h header file
which may not necessarily be installed on your system. Check the output
./configure to examine the configuration that ctools has
detected. You may see the following:
* Python (yes) * Python.h (yes) * swig (yes) * Python wrappers (yes)
Installing on Solaris
Although ctools build on Solaris using the Sun compiler, there are problems with global symbols in shared libraries and exception catching, which prevents the FITS interface to work correctly. ctools have however been built and tested successfully using the GNU compiler, and this is the only build method that is currently supported. Problems have also been encountered when compiling cfitsio versions more recent than 3.250. The problems have been reported to the cfitsio developer team, and are likely to be solved in the future. For the time being, it is recommended to use cfitsio version 3.250 on Solaris.
Installing on OpenSolaris
On OpenSolaris, the same problems concerning the SunStudio compiler occur as for Solaris, and also here, the GNU compiler is the recommended tool to build ctools. Also here, cfitsio version 3.250 is the recommended library as more recent version feature relocation problems. ctools have been tested using gcc 4.3.2 on OpenSolaris 2009.06. Make sure to create the symbolic links
$ ln -s /usr/bin/gcc4.3.2 /usr/bin/gcc $ ln -s /usr/bin/g++4.3.2 /usr/bin/g++
which are not there by default. This avoids warnings during compilation.
Conda installation on Ubuntu 18.04
According to this thread conda does not successfully work on Ubuntu 18.04 and consequently you will also encounter problems when installing ctools on Ubuntu 18.04 via conda. If you want to use conda on Ubuntu you need to upgrade to a newer version. Successful operations were reported, for example, for Ubuntu 20.04.
Binned analysis is biased for coarse binning
When performing a binned or stacked analysis you should make sure that the spatial and spectral binning is sufficiently fine grained. The spatial binning should be better than the best angular resolution over the energy range of interest. Use a typical value of 0.02 degrees per pixel for the spatial binning and 10 bins per decade for the spectral binning. If the binning is too coarse, the spectral parameters that are fitted will be biased.
Model fitting issues¶
Pivot energy should be comprised in energy range of fitted data
The pivot energy of a spectral model, such as for example a power law model, should be comprised within the energy range of the fitted data, otherwise some fit instabilities may occur.
Errors become unreliable when fitting the pivot energy
have a pivot energy, specified by the
Scale parameter, and this
pivot energy can not be determined in a fit together with the other
model parameters. The reason is that the pivot energy is not an
independent parameter of these models, and hence when all other
spectral parameters are free, the pivot energy is unconstrained.
So please make sure that the pivot energy is fixed, or fix other
parameters of the model to assure non-degeneracy of the free
Broken power law has unreliable errors
The broken power law spectral model has unreliable errors, specifically for the prefactor and the break value. Errors are in general too large, and this is related to the fact that the law’s gradient is discontinuous in energy. There is not very much we can do about it, it’s inherent in the law.
Shell model is biased when width is comparable to angular resolution
When the width of the shell model becomes comparable to or smaller than the angular resolution, the shell width tends to be overestimated while the shell radius tends to be underestimated. The fitted shell width and radius should thus not be overinterpreted when the width is close to the angular resolution of CTA.
Elliptical Gaussian model fit converges slowly
The convergence of the elliptical Gaussian model can be slow and in some situations requires of the order of 20 iterations before the fit terminates. Nevertheless, the numerical accuracy of the model fitting results are satisfactory.