Set GModelSpatialEllipticalGauss::theta_max() value to 2.5 times the maximum of …

…semimajor and semiminor axis. Before the limitation was 5 times that values and this led to convergence problems when fitting the model. Limiting the model to 2.5 times the larger axis seems to avoid the problem. Note that this is certainly not a perfect fix of the problem, but it helps to make the working model for now. Solves at least #1561.

Set GModelSpatialEllipticalGauss::theta_max() value to 2.5 times the maximum of …
…semimajor and semiminor axis. Before the limitation was 5 times that values and this led to convergence problems when fitting the model. Limiting the model to 2.5 times the larger axis seems to avoid the problem. Note that this is certainly not a perfect fix of the problem, but it helps to make the working model for now. Solves at least #1561.
Jürgen Knödlseder
1 parent b653863b
Showing 1 changed file with 9 additions and 6 deletions
src/model/GModelSpatialEllipticalGauss.cpp
@@ -404,7 +404,7 @@ GSkyDir GModelSpatialEllipticalGauss::mc(const GEnergy&amp; energy,
  
  
 /***********************************************************************//**
- * @brief Checks where model contains specified sky direction
+ * @brief Checks whether model contains specified sky direction
  *
  * @param[in] dir Sky direction.
  * @param[in] margin Margin to be added to sky direction (degrees)
@@ -431,16 +431,19 @@ bool GModelSpatialEllipticalGauss::contains(const GSkyDir&amp; dir,
  *
  * @return Returns maximum model radius.
  *
- * Returns \f$5 \sigma\f$ as approximate edge of the Gaussian. This limit
- * is of course arbitrary, but allows to limit the integration region for
- * response computation.
+ * Returns the maximum of \f$2.5\f$ semimajor() and \f$2.5\f$ semiminor() as
+ * approximate edge of the Gaussian. This limit is of course arbitrary, but
+ * allows to limit the integration region for response computation. The value
+ * of 2.5 has been determined by experiment. Ideally, a large value should
+ * be used to properly take into account the tails of the distribution, but
+ * this apparently leads to fit convergence problems (#1561).
  ***************************************************************************/
 double GModelSpatialEllipticalGauss::theta_max(void) const
 {
     // Set maximum model radius
     double theta_max = (semimajor() > semiminor())
-                       ? semimajor() * gammalib::deg2rad * 5.0
-                       : semiminor() * gammalib::deg2rad * 5.0;
+                       ? semimajor() * gammalib::deg2rad * 2.5
+                       : semiminor() * gammalib::deg2rad * 2.5;
  
     // Return value
     return theta_max;