Changes the default number of walks to 30x ndim (previously 10x ndims).

The default settings are insufficient for the pp-tests of the high-spin 128s basis to pass (#388 (closed)). This changes fixes that. The downside of this is it is setting out default settings by CBC inference. So people doing "easier" problems will need to tone down the settings.

bilby/core/sampler/dynesty.py

@@ -55,8 +55,8 @@ class Dynesty(NestedSampler):
         Method used to sample uniformly within the likelihood constraints,
         conditioned on the provided bounds
     walks: int
-        Number of walks taken if using `sample='rwalk'`, defaults to `ndim * 5`
-    dlogz: float, (0.1)
+        Number of walks taken if using `sample='rwalk'`, defaults to `ndim * 30`
+    dlogz: float, (0.5)
         Stopping criteria
     verbose: Bool
         If true, print information information about the convergence during
@@ -85,7 +85,7 @@ class Dynesty(NestedSampler):
                           enlarge=None, bootstrap=None, vol_dec=0.5, vol_check=2.0,
                           facc=0.5, slices=5,
                           update_interval=None, print_func=None,
-                          dlogz=0.1, maxiter=None, maxcall=None,
+                          dlogz=0.5, maxiter=None, maxcall=None,
                           logl_max=np.inf, add_live=True, print_progress=True,
                           save_bounds=False)
 
@@ -156,7 +156,7 @@ class Dynesty(NestedSampler):
 
     def _verify_kwargs_against_default_kwargs(self):
         if not self.kwargs['walks']:
-            self.kwargs['walks'] = self.ndim * 10
+            self.kwargs['walks'] = self.ndim * 30
         if not self.kwargs['update_interval']:
             self.kwargs['update_interval'] = int(0.6 * self.kwargs['nlive'])
         if not self.kwargs['print_func']: