Fix #475, add a corresponding test

also deprecate repeated run_nested() runs

CHANGELOG.md

@@ -16,9 +16,9 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 ### Changed
 - Get rid of npdim option that at some point may have allowed the prior transformation to return higher dimensional vector than the inputs. Note that due to this change, restoring the checkpoint from previous version of the dynesty won't be possible) (issues #456, #457) (original issue reported by @MichaelDAlbrow, fixed by @segasai )
 ### Fixed
-- Fix the way the additional arguments are treated when working with dynesty's pool. Previously those only could have been passed through dynesty.pool.Pool() constructor. Now they can still be provided directly to the sampler (not recommended) (reported by @eteq, fixed by @segasai )
-- change the .ptp() method to np.ptp() function as it is deprecated in numpy 2.0 (reported and patched by @joezuntz)
-
+- Fix the way the additional arguments are treated when working with dynesty's pool. Previously those only could have been passed through dynesty.pool.Pool() constructor. Now they can still be provided directly to the sampler (not recommended) ( #464 , reported by @eteq, fixed by @segasai )
+- change the .ptp() method to np.ptp() function as it is deprecated in numpy 2.0 ( #478 , reported and patched by @joezuntz)
+- Fix an error if you use run_nested() several times (i.e. with maxiter option) while using blob=True. ( #475 , reported by @carlosRmelo,)
 
 ## [2.1.3] - 2023-10-04
 ### Added

docs/source/index.rst

@@ -103,6 +103,7 @@ Changelog
 - Get rid of npdim option that at some point may have allowed the prior transformation to return higher dimensional vector than the inputs. Note that due to this change, restoring the checkpoint from previous version of the dynesty won't be possible) (issues #456, #457) (original issue reported by @MichaelDAlbrow, fixed by @segasai )
 - Fix the way the additional arguments are treated when working with dynesty's pool. Previously those only could have been passed through dynesty.pool.Pool() constructor. Now they can still be provided directly to the sampler (not recommended) (reported by @eteq, fixed by @segasai )
 - change the .ptp() method to np.ptp() function as it is deprecated in numpy 2.0 (reported and patched by @joezuntz)
+- Fix an error if you use run_nested() several times (i.e. with maxiter option) while using blob=True. ( #475 , reported by @carlosRmelo,)
 
 2.1.3 (2023-10-04)
 ------------------

py/dynesty/sampler.py

@@ -589,7 +589,7 @@ def _remove_live_points(self):
                 for k in [
                         'id', 'u', 'v', 'logl', 'logvol', 'logwt', 'logz',
                         'logzvar', 'h', 'nc', 'boundidx', 'it', 'bounditer',
-                        'scale'
+                        'scale', 'blob'
                 ]:
                     del self.saved_run[k][-self.nlive:]
         else:
@@ -729,6 +729,11 @@ def sample(self,
         else:
             # Remove live points (if added) from previous run.
             if self.added_live and not resume:
+                warnings.warn(
+                    'Repeatedly running sample() or run_nested() '
+                    '(when not just resuming an existing run is considered '
+                    'deprecated and will be removed in the future',
+                    DeprecationWarning)
                 self._remove_live_points()
 
             # Get final state from previous run.

tests/test_blob.py

@@ -85,6 +85,25 @@ def test_gaussian():
     assert np.all(res['blob'] == res['samples'])
 
 
+def test_restart():
+    rstate = get_rstate()
+    g = Gaussian()
+    sampler = 'rslice'  # doing this sampler for static
+    # unifor for dynamic
+    sampler = dynesty.NestedSampler(g.loglikelihood_with_blob,
+                                    g.prior_transform,
+                                    g.ndim,
+                                    nlive=nlive,
+                                    rstate=rstate,
+                                    sample=sampler,
+                                    blob=True)
+    sampler.run_nested(print_progress=printing, maxiter=100)
+    sampler.run_nested(print_progress=printing, maxiter=100)
+    res = sampler.results
+    assert res['blob'].shape == (len(res['samples']), 3)
+    assert np.all(res['blob'] == res['samples'])
+
+
 def test_gaussian_livepts():
     # test we can provide starting points while using blobs
     rstate = get_rstate()