Prequel to dependency upgrades

.github/workflows/test.yml

@@ -9,12 +9,14 @@ on:
       - main
     paths:
       - matminer/**
+      - requirements/**
 
   pull_request:
     branches:
       - main
     paths:
       - matminer/**
+      - requirements/**
 
   workflow_dispatch:
     inputs:

matminer/data_retrieval/retrieve_AFLOW.py

@@ -218,13 +218,11 @@ def _add_filters(self, pymongo_query):
         """
 
         for str_property, value in pymongo_query.items():
-
             # converts str representation of property to aflow.Keyword
             keyword = getattr(K, str_property)
 
             if isinstance(value, dict):  # handles special operators
                 for inner_key, inner_value in value.items():
-
                     if inner_key == "$in":
                         self.filter(
                             reduce(

matminer/data_retrieval/tests/test_retrieve_MPDS.py

@@ -36,7 +36,6 @@ def setUp(self):
     @unittest.skipIf(on_ci.upper() == "TRUE", "Bad Datasource-GHActions pipeline")
     @unittest.skipIf("MPDS_KEY" not in os.environ, "MPDS_KEY env var not set")
     def test_valid_answer(self):
-
         client = MPDSDataRetrieval()
         answer = client.get_data(self.test_request, fields={})
 

matminer/datasets/tests/test_datasets.py

@@ -21,7 +21,6 @@ def universal_dataset_check(
         bool_headers=None,
         test_func=None,
     ):
-
         # "Hard" integrity checks that take a long time.
         # These tests only run if the MATMINER_DATASET_FULL_TEST
         # environment variable is set to True
@@ -577,7 +576,6 @@ def test_expt_gap_kingsbury(self):
         self.universal_dataset_check("expt_gap_kingsbury", object_headers, numeric_headers)
 
     def test_expt_formation_enthalpy_kingsbury(self):
-
         object_headers = ["formula", "likely_mpid", "phaseinfo", "reference"]
 
         numeric_headers = ["expt_form_e", "uncertainty"]

matminer/datasets/utils.py

@@ -78,7 +78,6 @@ def _validate_dataset(data_path, url=None, file_hash=None, download_if_missing=T
     do_download = False
     # If the file doesn't exist, download it
     if not os.path.exists(data_path):
-
         # Ensure proper arguments for download
         if not download_if_missing:
             raise OSError("Data not found and download_if_missing set to False")

matminer/featurizers/conversions.py

@@ -643,7 +643,6 @@ class PymatgenFunctionApplicator(ConversionFeaturizer):
     """
 
     def __init__(self, func, func_args=None, func_kwargs=None, target_col_id=None, overwrite_data=False):
-
         if not callable(func):
             raise TypeError(f"Function {func} is not callable!")
 

matminer/featurizers/dos.py

@@ -597,7 +597,6 @@ def get_site_dos_scores(dos, idx, decay_length, sampling_resolution, gaussian_sm
     orbital_scores = {}
     proj = dos.get_site_spd_dos(site)
     for orb in proj:
-
         # smear dos for spin up and down
         smear_dos = proj[orb].get_smeared_densities(gaussian_smear)
         dos_up = smear_dos[Spin.up]

matminer/featurizers/function.py

@@ -63,7 +63,7 @@ class FunctionFeaturizer(BaseFeaturizer):
             include pairwise combined features
         postprocess (function or type): type to cast functional outputs
             to, if, for example, you want to include the possibility of
-            complex numbers in your outputs, use postprocess=np.complex,
+            complex numbers in your outputs, use postprocess=np.complex128,
             defaults to float
         combo_function (function): function to combine multi-features,
             defaults to np.prod (i.e. cumulative product of expressions),
@@ -84,7 +84,6 @@ def __init__(
         combo_function=None,
         latexify_labels=False,
     ):
-
         self.expressions = expressions or default_exps
         self.multi_feature_depth = multi_feature_depth
         self.combo_function = combo_function or np.prod

matminer/featurizers/site/chemical.py

@@ -99,6 +99,11 @@ def fit(self, X, y=None):
         Returns:
             self
         """
+        if isinstance(X, (list, tuple)):
+            # Required for numpy 1.24 due to changes in the way numpy casts
+            # object arrays.
+            X = np.array(X, dtype=object)
+
         structs = np.atleast_2d(X)[:, 0]
         if not all([isinstance(struct, Structure) for struct in structs]):
             raise TypeError("This fit requires an array-like input of Pymatgen " "Structures and sites!")

matminer/featurizers/site/external.py

@@ -148,9 +148,9 @@ def fit(self, X, y=None):
         self.atomic_numbers = elements
         self.soap = SOAP_dscribe(
             species=self.atomic_numbers,
-            rcut=self.rcut,
-            nmax=self.nmax,
-            lmax=self.lmax,
+            r_cut=self.rcut,
+            n_max=self.nmax,
+            l_max=self.lmax,
             sigma=self.sigma,
             rbf=self.rbf,
             periodic=self.periodic,
@@ -165,7 +165,7 @@ def fit(self, X, y=None):
     def featurize(self, struct, idx):
         self._check_fitted()
         s_ase = self.adaptor.get_atoms(struct)
-        return self.soap.create(s_ase, positions=[idx], n_jobs=self.n_jobs).tolist()[0]
+        return self.soap.create(s_ase, centers=[idx], n_jobs=self.n_jobs).tolist()[0]
 
     def feature_labels(self):
         self._check_fitted()

matminer/featurizers/site/fingerprint.py

@@ -167,7 +167,6 @@ def __init__(
         dist_exp=2,
         zero_ops=True,
     ):
-
         cn_target_motif_op = load_cn_target_motif_op()
         cn_motif_op_params = load_cn_motif_op_params()
 

matminer/featurizers/site/rdf.py

@@ -97,7 +97,6 @@ def g4(etas, zetas, gammas, neigh_dist, neigh_coords, cutoff):
 
         # Loop over each neighbor j
         for j, neigh_j in enumerate(neigh_coords):
-
             # Compute the distance of each neighbor (k) to r
             r_ij = neigh_dist[j]
             d_jk = neigh_coords[(j + 1) :] - neigh_coords[j]

matminer/featurizers/structure/bonding.py

@@ -407,7 +407,6 @@ def _approximate_bonds(self, local_bonds):
                 nearest = []
                 d_min = None
                 for abss in abonds_species.keys():
-
                     # The distance between bonds is euclidean. To get a good
                     # measure of the coordinate between mendeleev numbers for
                     # each specie, we use the minimum difference. ie, for
@@ -715,7 +714,6 @@ class GlobalInstabilityIndex(BaseFeaturizer):
     """
 
     def __init__(self, r_cut=4.0, disordered_pymatgen=False):
-
         bv = IUCrBondValenceData()
         self.bv_values = bv.params
         self.r_cut = r_cut

matminer/featurizers/structure/composite.py

@@ -51,7 +51,6 @@ def __init__(
         use_ddf=True,
         use_nn=True,
     ):
-
         self.use_cell = use_cell
         self.use_chem = use_chem
         self.use_chg = use_chg

matminer/featurizers/structure/matrix.py

@@ -335,7 +335,7 @@ def get_ohv(self, sp, period_tag):
             ohd[l][curr_shell[2]] = 1
             nume += curr_shell[2]
             shell_num += 1
-        my_ohv = np.zeros(self.size, np.int)
+        my_ohv = np.zeros(self.size, np.int32)
         k = 0
         for j in range(4):
             for i in range(2 * (2 * j + 1)):

matminer/featurizers/structure/rdf.py

@@ -340,7 +340,7 @@ def featurize(self, s):
         # Add oxidation states.
         struct = ValenceIonicRadiusEvaluator(struct).structure
 
-        distribution = np.zeros(self.nbins, dtype=np.float)
+        distribution = np.zeros(self.nbins, dtype=np.float64)
 
         for site in struct.sites:
             this_charge = float(site.specie.oxi_state)

matminer/featurizers/structure/sites.py

@@ -330,7 +330,6 @@ def featurize(self, s):
         return np.hstack(output)
 
     def compute_pssf(self, s, e):
-
         # This code is extremely similar to super().featurize(). The key
         # difference is that only one specific element is analyzed.
 

matminer/featurizers/structure/tests/test_bonding.py

@@ -19,7 +19,6 @@
 
 class BondingStructureTest(StructureFeaturesTest):
     def test_bondfractions(self):
-
         # Test individual structures with featurize
         bf_md = BondFractions.from_preset("MinimumDistanceNN")
         bf_md.no_oxi = True
@@ -62,7 +61,6 @@ def test_bondfractions(self):
         self.assertArrayEqual(df["Ni - Ni bond frac."].to_numpy(), [0.0, 0.5])
 
     def test_bob(self):
-
         # Test a single fit and featurization
         scm = SineCoulombMatrix(flatten=False)
         bob = BagofBonds(coulomb_matrix=scm, token=" - ")

matminer/featurizers/structure/tests/test_misc.py

@@ -37,7 +37,6 @@ def test_composition_features(self):
         self.assertEqual(comp.implementors(), f.implementors())
 
     def test_xrd_powderPattern(self):
-
         # default settings test
         xpp = XRDPowderPattern()
         pattern = xpp.featurize(self.diamond)

matminer/featurizers/tests/test_base.py

@@ -449,7 +449,6 @@ def test_ignore_errors(self):
         # multiindex or not, and interaction over entries/featurizers
 
         for mi, re, n, iter_entries in product([True, False], [True, False], [1, 2], [True, False]):
-
             mf = MultipleFeaturizer([self.multi, self.single], iterate_over_entries=iter_entries)
             # Make some test data that will cause errors
             data = pd.DataFrame({"x": ["a", 2, 3]})

matminer/featurizers/tests/test_dos.py

@@ -28,7 +28,6 @@ def setUp(self):
         self.nb3sn_df = pd.DataFrame({"dos": [nb3sn_dos]})
 
     def test_SiteDOS(self):
-
         dos = self.df["dos"][0]
 
         # ensure that both sites give same scores (expected behavior for si)

matminer/featurizers/tests/test_function.py

@@ -50,7 +50,7 @@ def test_featurize(self):
 
         # Test complex functionality
         expressions = ["sqrt(x)"]
-        ff = FunctionFeaturizer(expressions=expressions, postprocess=np.complex)
+        ff = FunctionFeaturizer(expressions=expressions, postprocess=np.complex128)
         new_df = ff.fit_featurize_dataframe(self.test_df, "a", inplace=False)
         self.assertEqual(new_df["sqrt(a)"][0], 1j)
 

matminer/featurizers/utils/grdf.py

@@ -88,7 +88,7 @@ def __call__(self, r_ij):
         return np.logical_and(
             np.greater_equal(r_ij, self.start),
             np.less(r_ij, self.start + self.width),
-            dtype=np.float,
+            dtype=np.float64,
         )
 
     def volume(self, cutoff):

matminer/utils/tests/test_io.py

@@ -52,7 +52,6 @@ def setUp(self):
         self.df = pd.DataFrame(data={"structure": [self.diamond]})
 
     def test_store_dataframe_as_json(self):
-
         # check write produces correct file
         temp_file = os.path.join(self.temp_folder, "test_dataframe.json")
         test_file = os.path.join(test_dir, "dataframe.json")
@@ -121,7 +120,6 @@ def test_store_dataframe_as_json(self):
         self.assertDictsAlmostEqual(temp_data, test_data)
 
     def test_load_dataframe_from_json(self):
-
         df = load_dataframe_from_json(os.path.join(test_dir, "dataframe.json"))
         self.assertTrue(self.diamond == df["structure"][0], "Dataframe contents do not match")
 

requirements/macos-latest_py3.10_extras.txt

@@ -58,7 +58,7 @@ docopt==0.6.2
     # via coveralls
 docutils==0.19
     # via sphinx
-dscribe==1.2.2
+dscribe==2.0.0
     # via matminer (setup.py)
 emmet-core==0.38.9
     # via mp-api

requirements/macos-latest_py3.8_extras.txt

@@ -58,7 +58,7 @@ docopt==0.6.2
     # via coveralls
 docutils==0.19
     # via sphinx
-dscribe==1.2.2
+dscribe==2.0.0
     # via matminer (setup.py)
 emmet-core==0.38.9
     # via mp-api

requirements/macos-latest_py3.9_extras.txt

@@ -58,7 +58,7 @@ docopt==0.6.2
     # via coveralls
 docutils==0.19
     # via sphinx
-dscribe==1.2.2
+dscribe==2.0.0
     # via matminer (setup.py)
 emmet-core==0.38.9
     # via mp-api

requirements/ubuntu-latest_py3.10_extras.txt

@@ -58,7 +58,7 @@ docopt==0.6.2
     # via coveralls
 docutils==0.19
     # via sphinx
-dscribe==1.2.2
+dscribe==2.0.0
     # via matminer (setup.py)
 emmet-core==0.38.9
     # via mp-api

requirements/ubuntu-latest_py3.8_extras.txt

@@ -58,7 +58,7 @@ docopt==0.6.2
     # via coveralls
 docutils==0.19
     # via sphinx
-dscribe==1.2.2
+dscribe==2.0.0
     # via matminer (setup.py)
 emmet-core==0.38.9
     # via mp-api

requirements/ubuntu-latest_py3.9_extras.txt

@@ -58,7 +58,7 @@ docopt==0.6.2
     # via coveralls
 docutils==0.19
     # via sphinx
-dscribe==1.2.2
+dscribe==2.0.0
     # via matminer (setup.py)
 emmet-core==0.38.9
     # via mp-api

setup.py

@@ -14,14 +14,26 @@ def local_version(version):
 
 extras_require = {
     "mpds": ["ujson", "jmespath", "httplib2", "ase", "jsonschema"],
-    "dscribe": ["dscribe"],
+    "dscribe": ["dscribe~=2.0"],
     "mdfforge": ["mdf-forge"],
     "aflow": ["aflow"],
     "citrine": ["citrination-client"],
-    "dev": ["pytest", "pytest-cov", "pytest-timeout", "coverage", "coveralls", "flake8", "black", "pylint", "sphinx",],
+    "dev": [
+        "pytest",
+        "pytest-cov",
+        "pytest-timeout",
+        "coverage",
+        "coveralls",
+        "flake8",
+        "black",
+        "pylint",
+        "sphinx",
+    ],
 }
 tests_require = [r for v in extras_require.values() for r in v]
 
+extras_require["tests"] = tests_require
+
 if __name__ == "__main__":
     setup(
         name="matminer",
@@ -44,7 +56,7 @@ def local_version(version):
         install_requires=[
             "numpy>=1.20.1",
             "requests",
-            "pandas",
+            "pandas~=1.5",
             "tqdm",
             "pymongo",
             "future",