Eliminate deprecation warnings from scipy and pandas (#2951)

openmc/data/photon.py

@@ -124,7 +124,7 @@ class AtomicRelaxation(EqualityMixin):
         Dictionary indicating the number of electrons in a subshell when neutral
         (values) for given subshells (keys). The subshells should be given as
         strings, e.g., 'K', 'L1', 'L2', etc.
-    transitions : pandas.DataFrame
+    transitions : dict of str to pandas.DataFrame
         Dictionary indicating allowed transitions and their probabilities
         (values) for given subshells (keys). The subshells should be given as
         strings, e.g., 'K', 'L1', 'L2', etc. The transitions are represented as
@@ -363,8 +363,9 @@ def from_hdf5(cls, group):
                 df = pd.DataFrame(sub_group['transitions'][()],
                                   columns=columns)
                 # Replace float indexes back to subshell strings
-                df[columns[:2]] = df[columns[:2]].replace(
-                              np.arange(float(len(_SUBSHELLS))), _SUBSHELLS)
+                with pd.option_context('future.no_silent_downcasting', True):
+                    df[columns[:2]] = df[columns[:2]].replace(
+                                np.arange(float(len(_SUBSHELLS))), _SUBSHELLS)
                 transitions[shell] = df
 
         return cls(binding_energy, num_electrons, transitions)
@@ -387,8 +388,9 @@ def to_hdf5(self, group, shell):
 
         # Write transition data with replacements
         if shell in self.transitions:
-            df = self.transitions[shell].replace(
-                 _SUBSHELLS, range(len(_SUBSHELLS)))
+            with pd.option_context('future.no_silent_downcasting', True):
+                df = self.transitions[shell].replace(
+                    _SUBSHELLS, range(len(_SUBSHELLS)))
             group.create_dataset('transitions', data=df.values.astype(float))
 
 

openmc/mgxs_library.py

@@ -3,7 +3,7 @@
 
 import h5py
 import numpy as np
-from scipy.integrate import simps
+import scipy.integrate
 from scipy.interpolate import interp1d
 from scipy.special import eval_legendre
 
@@ -1823,6 +1823,12 @@ def convert_scatter_format(self, target_format, target_order=None):
         # Reset and re-generate XSdata.xs_shapes with the new scattering format
         xsdata._xs_shapes = None
 
+        # scipy 1.11+ prefers 'simpson', whereas older versions use 'simps'
+        if hasattr(scipy.integrate, 'simpson'):
+            integrate = scipy.integrate.simpson
+        else:
+            integrate = scipy.integrate.simps
+
         for i, temp in enumerate(xsdata.temperatures):
             orig_data = self._scatter_matrix[i]
             new_shape = orig_data.shape[:-1] + (xsdata.num_orders,)
@@ -1860,7 +1866,7 @@ def convert_scatter_format(self, target_format, target_order=None):
                                 table_fine[..., imu] += ((l + 0.5)
                                      * eval_legendre(l, mu_fine[imu]) *
                                      orig_data[..., l])
-                        new_data[..., h_bin] = simps(table_fine, mu_fine)
+                        new_data[..., h_bin] = integrate(table_fine, mu_fine)
 
             elif self.scatter_format == SCATTER_TABULAR:
                 # Calculate the mu points of the current data
@@ -1874,7 +1880,7 @@ def convert_scatter_format(self, target_format, target_order=None):
                     for l in range(xsdata.num_orders):
                         y = (interp1d(mu_self, orig_data)(mu_fine) *
                              eval_legendre(l, mu_fine))
-                        new_data[..., l] = simps(y, mu_fine)
+                        new_data[..., l] = integrate(y, mu_fine)
 
                 elif target_format == SCATTER_TABULAR:
                     # Simply use an interpolating function to get the new data
@@ -1893,7 +1899,7 @@ def convert_scatter_format(self, target_format, target_order=None):
                     interp = interp1d(mu_self, orig_data)
                     for h_bin in range(xsdata.num_orders):
                         mu_fine = np.linspace(mu[h_bin], mu[h_bin + 1], _NMU)
-                        new_data[..., h_bin] = simps(interp(mu_fine), mu_fine)
+                        new_data[..., h_bin] = integrate(interp(mu_fine), mu_fine)
 
             elif self.scatter_format == SCATTER_HISTOGRAM:
                 # The histogram format does not have enough information to
@@ -1919,7 +1925,7 @@ def convert_scatter_format(self, target_format, target_order=None):
                     mu_fine = np.linspace(-1, 1, _NMU)
                     for l in range(xsdata.num_orders):
                         y = interp(mu_fine) * norm * eval_legendre(l, mu_fine)
-                        new_data[..., l] = simps(y, mu_fine)
+                        new_data[..., l] = integrate(y, mu_fine)
 
                 elif target_format == SCATTER_TABULAR:
                     # Simply use an interpolating function to get the new data
@@ -1938,7 +1944,7 @@ def convert_scatter_format(self, target_format, target_order=None):
                     for h_bin in range(xsdata.num_orders):
                         mu_fine = np.linspace(mu[h_bin], mu[h_bin + 1], _NMU)
                         new_data[..., h_bin] = \
-                            norm * simps(interp(mu_fine), mu_fine)
+                            norm * integrate(interp(mu_fine), mu_fine)
 
             # Remove small values resulting from numerical precision issues
             new_data[..., np.abs(new_data) < 1.E-10] = 0.

tests/unit_tests/test_data_photon.py

@@ -1,5 +1,3 @@
-#!/usr/bin/env python
-
 from collections.abc import Mapping, Callable
 import os
 from pathlib import Path
@@ -123,6 +121,8 @@ def test_reactions(element, reaction):
         reactions[18]
 
 
+# TODO: Remove skip when support is Python 3.9+
+@pytest.mark.skipif(not hasattr(pd.options, 'future'), reason='pandas version too old')
 @pytest.mark.parametrize('element', ['Pu'], indirect=True)
 def test_export_to_hdf5(tmpdir, element):
     filename = str(tmpdir.join('tmp.h5'))
@@ -146,8 +146,9 @@ def test_export_to_hdf5(tmpdir, element):
     # Export to hdf5 again
     element2.export_to_hdf5(filename, 'w')
 
+
 def test_photodat_only(run_in_tmpdir):
     endf_dir = Path(os.environ['OPENMC_ENDF_DATA'])
     photoatomic_file = endf_dir / 'photoat' / 'photoat-001_H_000.endf'
     data = openmc.data.IncidentPhoton.from_endf(photoatomic_file)
-    data.export_to_hdf5('tmp.h5', 'w')
+    data.export_to_hdf5('tmp.h5', 'w')