CP2K tweaks (#1056)

* refactor dict merges to use pipe operator

plus typo fix

* more readable var names

* Cp2kInputSet.is_valid warn that not implemented

* Cp2kInputGenerator.get_input_set allow optional_files=False, meaning no optional files will be written

before there was no way to disable writing potentially redundant BASIS and POTENTIAL files

* don't write kpoints if user_kpoints_settings or base KPOINTS config is None

KPOINTS are not compatible with orbital transformation (OT) mode and CP2K will crash if found

* add asserts in cp2k/test_drones.py test_structure_optimization

* fix test_structure_optimization assert doc.completed_at

* more unit tests in cp2k/test_files.py

* more atomate2.cp2k.powerups unit tests

Author: janosh

.pre-commit-config.yaml

@@ -3,7 +3,7 @@ default_language_version:
 exclude: ^(.github/|tests/test_data/abinit/)
 repos:
 - repo: https://github.com/charliermarsh/ruff-pre-commit
-  rev: v0.6.9
+  rev: v0.7.3
   hooks:
   - id: ruff
     args: [--fix]
@@ -17,7 +17,7 @@ repos:
   - id: end-of-file-fixer
   - id: trailing-whitespace
 - repo: https://github.com/asottile/blacken-docs
-  rev: 1.18.0
+  rev: 1.19.1
   hooks:
   - id: blacken-docs
     additional_dependencies: [black]
@@ -30,7 +30,7 @@ repos:
   - id: rst-directive-colons
   - id: rst-inline-touching-normal
 - repo: https://github.com/pre-commit/mirrors-mypy
-  rev: v1.11.2
+  rev: v1.13.0
   hooks:
   - id: mypy
     files: ^src/
@@ -45,7 +45,7 @@ repos:
     args: [--ignore-words-list, 'titel,statics,ba,nd,te,atomate']
     types_or: [python, rst, markdown]
 - repo: https://github.com/kynan/nbstripout
-  rev: 0.7.1
+  rev: 0.8.0
   hooks:
     - id: nbstripout
       args:

pyproject.toml

@@ -34,28 +34,28 @@ dependencies = [
     "numpy",
     "pydantic-settings>=2.0.3",
     "pydantic>=2.0.1",
-    "pymatgen>=2024.10.3",
+    "pymatgen>=2024.11.13",
 ]
 
 [project.optional-dependencies]
 abinit = ["abipy>=0.9.3"]
 amset = ["amset>=0.4.15", "pydash"]
 cclib = ["cclib"]
 mp = ["mp-api>=0.37.5"]
-phonons = ["phonopy>=1.10.8", "seekpath"]
+phonons = ["phonopy>=1.10.8", "seekpath>=2.0.0"]
 lobster = ["ijson>=3.2.2", "lobsterpy>=0.4.0"]
 defects = [
     "dscribe>=1.2.0",
     "pymatgen-analysis-defects>=2024.5.11",
-    "python-ulid",
+    "python-ulid>=2.7",
 ]
 forcefields = [
     "ase>=3.23.0",
     "calorine<=2.2.1",
     "chgnet>=0.2.2",
     "mace-torch>=0.3.3",
-    "torchdata<=0.7.1",
     "matgl>=1.1.3",
+    "torchdata<=0.7.1",
     # quippy-ase support for py3.12 tracked in https://github.com/libAtoms/QUIP/issues/645
     "quippy-ase>=0.9.14; python_version < '3.12'",
     "sevenn>=0.9.3",
@@ -112,7 +112,7 @@ strict = [
     "pydantic-settings==2.6.1",
     "pydantic==2.9.2",
     "pymatgen-analysis-defects==2024.7.19",
-    "pymatgen==2024.10.3",
+    "pymatgen==2024.11.13",
     "python-ulid==3.0.0",
     "seekpath==2.1.0",
     "tblite==0.3.0; python_version < '3.12'",

src/atomate2/common/jobs/mpmorph.py

@@ -73,7 +73,7 @@ def get_average_volume_from_mp_api(
     Returns
     -------
     float
-        The average volume per atom for the composition.
+        The average volume per atom for the composition in Angstrom^3.
     """
     from mp_api.client import MPRester
 

src/atomate2/common/schemas/magnetism.py

@@ -81,7 +81,7 @@ class MagneticOrderingRelaxation(BaseModel):
         None, description="Total magnetization normalized to per formula unit."
     )
     total_magnetization_per_unit_volume: Optional[float] = Field(
-        None, description="Total magnetiation normalized to per unit volume."
+        None, description="Total magnetization normalized to per unit volume."
     )
 
     @classmethod
@@ -175,7 +175,7 @@ class MagneticOrderingOutput(BaseModel):
         None, description="Total magnetization normalized to per formula unit."
     )
     total_magnetization_per_unit_volume: Optional[float] = Field(
-        None, description="Total magnetiation normalized to per unit volume."
+        None, description="Total magnetization normalized to per unit volume."
     )
     ordering_changed: Optional[bool] = Field(
         None,

src/atomate2/cp2k/files.py

@@ -65,22 +65,24 @@ def copy_cp2k_outputs(
     additional_cp2k_files = additional_cp2k_files or []
 
     # find required files
-    o = Cp2kOutput(src_dir / get_zfile(directory_listing, "cp2k.out"), auto_load=False)
-    o.parse_files()
+    cp2k_output = Cp2kOutput(
+        src_dir / get_zfile(directory_listing, "cp2k.out"), auto_load=False
+    )
+    cp2k_output.parse_files()
     if restart_to_input:
         additional_cp2k_files += ("restart",)
 
     # copy files
     additional_cp2k_files += ("wfn",)
     files = ["cp2k.inp", "cp2k.out"]
-    for f in set(additional_cp2k_files):
-        if f in o.filenames and o.filenames.get(f):
-            if isinstance(o.filenames[f], str):
-                files.append(Path(o.filenames[f]).name)
+    for file in set(additional_cp2k_files):
+        if file in cp2k_output.filenames and cp2k_output.filenames.get(file):
+            if isinstance(cp2k_output.filenames[file], str):
+                files.append(Path(cp2k_output.filenames[file]).name)
             else:
-                files.append(Path(o.filenames[f][-1]).name)
+                files.append(Path(cp2k_output.filenames[file][-1]).name)
         else:
-            files.append(Path(f).name)
+            files.append(Path(file).name)
     all_files = [
         get_zfile(directory_listing, r + relax_ext, allow_missing=True) for r in files
     ]

src/atomate2/cp2k/flows/core.py

@@ -222,7 +222,7 @@ class HybridFlowMaker(Maker):
     initialize_with_pbe
         Whether or not to attach a pre-hybrid flow that can be used to
         kickstart the hybrid flow. This is treated differently than just
-        stiching flows together, because of the screening done in
+        stitching flows together, because of the screening done in
         __post_init__
     pbe_maker
         Maker for the initialization

src/atomate2/cp2k/run.py

@@ -112,7 +112,7 @@ def run_cp2k(
     else:
         raise ValueError(f"Unsupported {job_type=}")
 
-    c = Custodian(
+    custodian = Custodian(
         handlers,
         jobs,
         validators=validators,
@@ -122,7 +122,7 @@ def run_cp2k(
     )
 
     logger.info("Running CP2K using custodian.")
-    c.run()
+    custodian.run()
 
 
 def should_stop_children(

src/atomate2/cp2k/schemas/calculation.py

@@ -480,12 +480,14 @@ def _get_basis_and_potential_files(dir_name: Path) -> dict[Cp2kObject, DataFile]
     the basis/potential contained in these files.
     """
     data: dict[Cp2kObject, DataFile] = {}
-    if Path.exists(dir_name / "BASIS"):
-        data[Cp2kObject.BASIS] = BasisFile.from_file(str(dir_name / "BASIS"))  # type: ignore[index]
-    if Path.exists(dir_name / "POTENTIAL"):
-        data[Cp2kObject.POTENTIAL] = PotentialFile.from_file(  # type: ignore[index]
-            str(dir_name / "POTENTIAL")
-        )
+    for filename, cls, cp2k_object in (
+        (dir_name / "BASIS", BasisFile, Cp2kObject.BASIS),
+        (dir_name / "POTENTIAL", PotentialFile, Cp2kObject.POTENTIAL),
+    ):
+        if filename.exists():
+            content = filename.read_text().strip()
+            if content not in ("None", ""):  # ignore empty files
+                data[cp2k_object] = cls.from_str(content)  # type: ignore[index]
     return data
 
 

src/atomate2/cp2k/schemas/task.py

@@ -211,9 +211,14 @@ def from_cp2k_calc_doc(cls, calc_doc: Calculation) -> Self:
         OutputSummary
             The calculation output summary.
         """
-        if calc_doc.output.ionic_steps:
-            forces = calc_doc.output.ionic_steps[-1].get("forces")
-            stress = calc_doc.output.ionic_steps[-1].get("stress")
+        if calc_doc.output.ionic_steps:  # also handles for static calculations
+            final_step = calc_doc.output.ionic_steps[-1]
+            forces = final_step.get("forces")
+            # 2024-11-14 @janosh this method used to read "stress" from final_step
+            # (still read as fallback). CP2K docs don't mention "stress", only
+            # "stress_tensor". Unclear if this was a breaking change in CP2K or
+            # should always have been stress_tensor in this code.
+            stress = final_step.get("stress_tensor", final_step.get("stress"))
         else:
             forces = None
             stress = None
@@ -338,7 +343,7 @@ def from_directory(
         task_files = _find_cp2k_files(dir_name, volumetric_files=volumetric_files)
 
         if len(task_files) == 0:
-            raise FileNotFoundError("No CP2K files found!")
+            raise FileNotFoundError(f"No CP2K files found in {dir_name}")
 
         calcs_reversed = []
         all_cp2k_objects = []

src/atomate2/cp2k/sets/base.py

@@ -3,11 +3,12 @@
 from __future__ import annotations
 
 import os
+import warnings
 from copy import deepcopy
 from dataclasses import dataclass, field
 from importlib.resources import files as get_mod_path
 from pathlib import Path
-from typing import Any
+from typing import Any, Literal
 
 import numpy as np
 from monty.io import zopen
@@ -39,7 +40,7 @@ class Cp2kInputSet(InputSet):
     def __init__(
         self,
         cp2k_input: Cp2kInput,
-        optional_files: dict | None = None,
+        optional_files: dict | None | Literal[False] = None,
     ) -> None:
         """Initialize the set.
 
@@ -144,6 +145,10 @@ def from_directory(
     @property
     def is_valid(self) -> bool:
         """Whether the input set is valid."""
+        warnings.warn(
+            "Cp2kInputSet.is_valid is not yet implemented and will always return True.",
+            stacklevel=2,
+        )
         return True
 
 
@@ -182,7 +187,7 @@ def get_input_set(
         self,
         structure: Structure | Molecule = None,
         prev_dir: str | Path = None,
-        optional_files: dict | None = None,
+        optional_files: dict | None | Literal[False] = None,
     ) -> Cp2kInputSet:
         """Get a CP2K input set.
 
@@ -197,6 +202,8 @@ def get_input_set(
             A previous directory to generate the input set from.
         optional_files
             Additional files (e.g. vdw kernel file) to be included in the input set.
+            If False, no optional files will be included. Defaults to writing a BASIS
+            and POTENTIAL file.
 
         Returns
         -------
@@ -224,15 +231,16 @@ def get_input_set(
             prev_input,
             input_updates,
         )
-        optional_files = optional_files or {}
-        optional_files["basis"] = {
-            "filename": "BASIS",
-            "object": self._get_basis_file(cp2k_input=cp2k_input),
-        }
-        optional_files["potential"] = {
-            "filename": "POTENTIAL",
-            "object": self._get_potential_file(cp2k_input=cp2k_input),
-        }
+        if optional_files is not False:
+            optional_files = optional_files or {}
+            optional_files["basis"] = {
+                "filename": "BASIS",
+                "object": self._get_basis_file(cp2k_input=cp2k_input),
+            }
+            optional_files["potential"] = {
+                "filename": "POTENTIAL",
+                "object": self._get_potential_file(cp2k_input=cp2k_input),
+            }
 
         return Cp2kInputSet(cp2k_input=cp2k_input, optional_files=optional_files)
 
@@ -323,11 +331,7 @@ def _get_input(
         # Generate base input but override with user input settings
         input_settings = recursive_update(input_settings, input_updates)
         input_settings = recursive_update(input_settings, self.user_input_settings)
-        overrides = (
-            input_settings.pop("override_default_params")
-            if "override_default_params" in input_settings
-            else {}
-        )
+        overrides = input_settings.pop("override_default_params", {})
         cp2k_input = DftSet(structure=structure, kpoints=kpoints, **input_settings)
 
         for setting in input_settings:
@@ -386,6 +390,15 @@ def _get_kpoints(
         """Get the kpoints object."""
         kpoints_updates = kpoints_updates or {}
 
+        # don't write kpoints if user_kpoints_settings or base KPOINTS config is None
+        # KPOINTS are not compatible with orbital transformation mode and CP2K will
+        # crash if found
+        if (
+            self.user_kpoints_settings is None
+            or self.config_dict.get("KPOINTS") is None
+        ):
+            return None
+
         # use user setting if set otherwise default to base config settings
         if self.user_kpoints_settings != {}:
             kpt_config = deepcopy(self.user_kpoints_settings)