Update tests

doped/analysis.py

@@ -595,7 +595,10 @@ def __init__(
         self.error_tolerance = error_tolerance
         self.bulk_path = bulk_path
         self.subfolder = subfolder
-        self.bulk_band_gap_vr = bulk_band_gap_vr
+        if bulk_band_gap_vr and not isinstance(bulk_band_gap_vr, Vasprun):
+            self.bulk_band_gap_vr = get_vasprun(bulk_band_gap_vr, parse_projected_eigen=False)
+        else:
+            self.bulk_band_gap_vr = bulk_band_gap_vr
         self.processes = processes
         self.json_filename = json_filename
         self.parse_projected_eigen = parse_projected_eigen

doped/thermodynamics.py

@@ -1357,10 +1357,7 @@ def _parse_fermi_dos(
             fdos_band_gap = fdos.get_gap(tol=1e-4, abs_tol=True)  # rounds the DOS outputs to 4 dp
 
         if isinstance(bulk_dos, str):
-            if hasattr(self, "_fermi_dos_hash") and calculate_md5(bulk_dos) == self._fermi_dos_hash:
-                return self.fermi_dos  # already parsed, don't waste time re-parsing, & no need to re-warn
-
-            bulk_dos = get_vasprun(bulk_dos, parse_dos=True)
+            bulk_dos = get_vasprun(bulk_dos, parse_dos=True)  # converted to fdos in next block
 
         if isinstance(bulk_dos, Vasprun):  # either supplied Vasprun or parsed from string there
             fdos_band_gap, _cbm, fdos_vbm, _ = bulk_dos.eigenvalue_band_properties
@@ -1500,8 +1497,6 @@ def get_equilibrium_fermi_level(
         """
         if bulk_dos is not None:
             self.fermi_dos = self._parse_fermi_dos(bulk_dos, skip_check=skip_check)
-            if isinstance(bulk_dos, str):
-                self._fermi_dos_hash = calculate_md5(bulk_dos)  # to avoid re-parsing unnecessarily
         elif not hasattr(self, "fermi_dos"):
             raise ValueError(
                 "No bulk DOS calculation (`bulk_dos`) provided or previously parsed to "
@@ -1721,8 +1716,6 @@ def get_quenched_fermi_level_and_concentrations(
 
         if bulk_dos is not None:
             self.fermi_dos = self._parse_fermi_dos(bulk_dos, skip_check=kwargs.get("skip_check", False))
-            if isinstance(bulk_dos, str):
-                self._fermi_dos_hash = calculate_md5(bulk_dos)  # to avoid re-parsing unnecessarily
         elif not hasattr(self, "fermi_dos"):
             raise ValueError(
                 "No bulk DOS calculation (`bulk_dos`) provided or previously parsed to "
@@ -1767,6 +1760,8 @@ def get_quenched_fermi_level_and_concentrations(
                 # gap not 0
             )
             assert not isinstance(annealing_fermi_level, tuple)  # float w/ return_concs=False, for typing
+            self.fermi_dos = orig_fermi_dos  # reset to original DOS for quenched calculations
+
             annealing_defect_concentrations = self.get_equilibrium_concentrations(
                 chempots=chempots,
                 limit=limit,

tests/test_analysis.py

@@ -29,6 +29,7 @@
 from doped.generation import DefectsGenerator, get_defect_name_from_defect, get_defect_name_from_entry
 from doped.utils.eigenvalues import get_eigenvalue_analysis
 from doped.utils.parsing import (
+    Vasprun,
     get_defect_site_idxs_and_unrelaxed_structure,
     get_defect_type_and_composition_diff,
     get_orientational_degeneracy,
@@ -223,7 +224,7 @@ def _check_default_CdTe_DefectsParser_outputs(
             assert CdTe_dp.error_tolerance == 0.05
             assert CdTe_dp.bulk_path == self.CdTe_BULK_DATA_DIR  # automatically determined
             assert CdTe_dp.subfolder == "vasp_ncl"  # automatically determined
-            assert CdTe_dp.bulk_band_gap_path is None
+            assert CdTe_dp.bulk_band_gap_vr is None
 
         self._check_DefectsParser(CdTe_dp)
         assert (
@@ -446,7 +447,7 @@ def test_DefectsParser_CdTe_custom_settings(self):
                 dielectric=[9.13, 9.13, 9.13],
                 error_tolerance=0.01,
                 skip_corrections=False,
-                bulk_band_gap_path=self.CdTe_BULK_DATA_DIR,
+                bulk_band_gap_vr=f"{self.CdTe_BULK_DATA_DIR}/vasprun.xml",
                 processes=4,
                 json_filename="test_pop.json",
             )
@@ -476,7 +477,7 @@ def test_DefectsParser_CdTe_custom_settings(self):
         assert dp.output_path == self.CdTe_EXAMPLE_DIR
         assert dp.dielectric == [9.13, 9.13, 9.13]
         assert dp.error_tolerance == 0.01
-        assert dp.bulk_band_gap_path == self.CdTe_BULK_DATA_DIR
+        assert isinstance(dp.bulk_band_gap_vr, Vasprun)
         assert dp.processes == 4
         assert dp.json_filename == "test_pop.json"
 
@@ -1058,7 +1059,7 @@ def test_SrTiO3_diff_ISYM_bulk_defect_and_concentration_funcs(self):
 
         # test get_quenched_fermi_level_and_concentrations
         fermi_level, e_conc, h_conc, conc_df = thermo.get_quenched_fermi_level_and_concentrations(
-            bulk_dos_vr=f"{self.SrTiO3_DATA_DIR}/bulk_sp333/vasprun.xml",
+            bulk_dos=f"{self.SrTiO3_DATA_DIR}/bulk_sp333/vasprun.xml",
             annealing_temperature=300,
             per_charge=False,
         )
@@ -1072,7 +1073,7 @@ def test_SrTiO3_diff_ISYM_bulk_defect_and_concentration_funcs(self):
         assert conc_df.index.name == "Defect"
 
         fermi_level, e_conc, h_conc, conc_df = thermo.get_quenched_fermi_level_and_concentrations(
-            bulk_dos_vr=f"{self.SrTiO3_DATA_DIR}/bulk_sp333/vasprun.xml",
+            bulk_dos=f"{self.SrTiO3_DATA_DIR}/bulk_sp333/vasprun.xml",
             annealing_temperature=300,
             skip_formatting=True,
         )
@@ -1091,7 +1092,7 @@ def test_SrTiO3_diff_ISYM_bulk_defect_and_concentration_funcs(self):
             assert list(conc_df.iloc[i]) == row
 
         fermi_level, e_conc, h_conc, conc_df = thermo.get_quenched_fermi_level_and_concentrations(
-            bulk_dos_vr=f"{self.SrTiO3_DATA_DIR}/bulk_sp333/vasprun.xml",
+            bulk_dos=f"{self.SrTiO3_DATA_DIR}/bulk_sp333/vasprun.xml",
             annealing_temperature=300,
             per_site=True,
         )
@@ -1110,7 +1111,7 @@ def test_SrTiO3_diff_ISYM_bulk_defect_and_concentration_funcs(self):
             assert list(conc_df.iloc[i]) == row
 
         fermi_level, e_conc, h_conc, conc_df = thermo.get_quenched_fermi_level_and_concentrations(
-            bulk_dos_vr=f"{self.SrTiO3_DATA_DIR}/bulk_sp333/vasprun.xml",
+            bulk_dos=f"{self.SrTiO3_DATA_DIR}/bulk_sp333/vasprun.xml",
             annealing_temperature=300,
             per_site=True,
             per_charge=False,

tests/test_thermodynamics.py

@@ -20,7 +20,7 @@
 from monty.serialization import dumpfn, loadfn
 
 from doped.generation import _sort_defect_entries
-from doped.thermodynamics import DefectThermodynamics, scissor_dos
+from doped.thermodynamics import DefectThermodynamics, get_fermi_dos, scissor_dos
 from doped.utils.parsing import get_vasprun
 from doped.utils.symmetry import _get_sga, point_symmetry
 
@@ -321,13 +321,13 @@ def _check_defect_thermo(
         print([str(warning.message) for warning in w])  # for debugging
         assert not w
 
-        with pytest.raises(TypeError) as exc:
+        with pytest.raises(ValueError) as exc:
             defect_thermo.get_equilibrium_fermi_level()
-        assert "missing 1 required positional argument: 'bulk_dos_vr'" in str(exc.value)
+        assert "No bulk DOS calculation (`bulk_dos`) provided" in str(exc.value)
 
-        with pytest.raises(TypeError) as exc:
+        with pytest.raises(ValueError) as exc:
             defect_thermo.get_quenched_fermi_level_and_concentrations()
-        assert "missing 1 required positional argument: 'bulk_dos_vr'" in str(exc.value)
+        assert "No bulk DOS calculation (`bulk_dos`) provided" in str(exc.value)
 
         if defect_thermo.chempots is None:
             with pytest.raises(ValueError) as exc:
@@ -1956,10 +1956,11 @@ def setUpClass(cls):
         cls.defect_thermo = DefectThermodynamics(cls.defect_dict)
         cls.defect_thermo.chempots = cls.CdTe_chempots
 
-        cls.fermi_dos = loadfn(
-            os.path.join(cls.CdTe_EXAMPLE_DIR, "CdTe_prim_k181818_NKRED_2_fermi_dos.json")
+        cls.fermi_dos = get_fermi_dos(
+            os.path.join(cls.CdTe_EXAMPLE_DIR, "CdTe_prim_k181818_NKRED_2_vasprun.xml.gz")
         )
         cls.anneal_temperatures = np.arange(200, 1401, 50)
+        cls.reduced_anneal_temperatures = np.arange(200, 1401, 100)  # for quicker testing
 
         cls.annealing_dict = {}
 
@@ -2012,6 +2013,7 @@ def test_calculated_fermi_levels(self):
         )
         quenched_fermi_levels = np.array([v["fermi_level"] for k, v in self.annealing_dict.items()])
         assert np.isclose(np.mean(self.anneal_temperatures[12:16]), 875)
+        # remember this is LZ thermo, not FNV thermo shown in thermodynamics tutorial
         assert np.isclose(np.mean(quenched_fermi_levels[12:16]), 0.318674, atol=1e-3)
 
         ax.plot(
@@ -2094,10 +2096,10 @@ def test_calculated_fermi_level_k10(self):
         """
         k10_dos_vr_path = os.path.join(self.module_path, "data/CdTe/CdTe_prim_k101010_dos_vr.xml.gz")
 
-        for i, bulk_dos_vr in enumerate([k10_dos_vr_path, get_vasprun(k10_dos_vr_path, parse_dos=True)]):
+        for i, bulk_dos in enumerate([k10_dos_vr_path, get_vasprun(k10_dos_vr_path, parse_dos=True)]):
             print(f"Testing k10 DOS with thermo for {'str input' if i == 0 else 'DOS object input'}")
             quenched_fermi_levels = []
-            for anneal_temp in self.anneal_temperatures:
+            for anneal_temp in self.reduced_anneal_temperatures:
                 gap_shift = belas_linear_fit(anneal_temp) - 1.5
                 (
                     fermi_level,
@@ -2106,15 +2108,16 @@ def test_calculated_fermi_level_k10(self):
                     conc_df,
                 ) = self.defect_thermo.get_quenched_fermi_level_and_concentrations(
                     # quenching to 300K (default)
-                    bulk_dos_vr=bulk_dos_vr,
+                    bulk_dos=bulk_dos,
                     limit="Te-rich",
                     annealing_temperature=anneal_temp,
                     delta_gap=gap_shift,
                 )
                 quenched_fermi_levels += [fermi_level]
 
-            # (approx) same result as with k181818 NKRED=2 (0.319104 eV with this dos)
-            assert np.isclose(np.mean(quenched_fermi_levels[12:16]), 0.318674, atol=1e-3)
+            # (approx) same result as with k181818 NKRED=2 (0.31825 eV with this DOS)
+            # remember this is LZ thermo, not FNV thermo shown in thermodynamics tutorial
+            assert np.isclose(np.mean(quenched_fermi_levels[6:8]), 0.31825, atol=1e-3)
 
     @custom_mpl_image_compare(filename="CdTe_LZ_Te_rich_concentrations.png")
     def test_calculated_concentrations(self):