NonConvergingErrorHandler document AMIX maybe ineffective in helping SCF convergence (#358)

Author: janosh

.pre-commit-config.yaml

@@ -2,7 +2,7 @@ exclude: ^(docs|tests/files|tasks.py)
 
 ci:
   autoupdate_schedule: monthly
-  skip: [ mypy, pyright ]
+  skip: [mypy, pyright]
   autofix_commit_msg: pre-commit auto-fixes
   autoupdate_commit_msg: pre-commit autoupdate
 
@@ -11,7 +11,7 @@ repos:
     rev: v0.11.4
     hooks:
       - id: ruff
-        args: [ --fix, --unsafe-fixes ]
+        args: [--fix, --unsafe-fixes]
       - id: ruff-format
 
   - repo: https://github.com/pre-commit/pre-commit-hooks

src/custodian/vasp/handlers.py

@@ -1591,6 +1591,9 @@ class NonConvergingErrorHandler(ErrorHandler):
     Check if a run is hitting the maximum number of electronic steps at the
     last nionic_steps ionic steps (default=10). If so, change ALGO using a
     multi-step ladder scheme or kill the job.
+
+    In some cases (ALGO=All or ALGO=Normal and ISMEAR < 0), this handler also changes AMIX
+    and BMIX but unsure if this helps much. Some anecdotal evidence suggests it doesn't.
     """
 
     is_monitor = True

src/custodian/vasp/jobs.py

@@ -724,7 +724,7 @@ def terminate(self, directory="./") -> None:
                         proc.kill()
                         return
             except (psutil.NoSuchProcess, psutil.AccessDenied) as exc:
-                logger.warning(f"Exception {exc} encountered while killing VASP.")
+                logger.exception(f"Exception {exc} encountered while killing VASP.")
                 continue
 
         logger.warning(

src/custodian/vasp/utils.py

@@ -18,6 +18,7 @@ def _estimate_num_k_points_from_kspacing(structure: Structure, kspacing: float)
     Inputs:
         structure (Structure): structure used in the calculation
         kspacing (float): KSPACING used in the calculation
+
     Returns:
         tuple[int,int,int] : the number of estimated k-points on each axis.