Add simple caching for dispersions

Author: MarJMue

src/elli/dispersions/base_dispersion.py

@@ -8,6 +8,7 @@
 import numpy as np
 import numpy.typing as npt
 import pandas as pd
+from lmfit import Parameter
 from numpy.lib.scimath import sqrt
 
 from .. import dispersions
@@ -44,6 +45,14 @@ def _guard_invalid_params(params1, params2):
             missing_param_strings = ", ".join(f"{p}" for p in missing_params)
             raise InvalidParameters(f"Invalid parameter(s): {missing_param_strings}")
 
+    @staticmethod
+    def _hash_params(params: dict | list[dict]) -> int:
+        """Creates an single_params_dict or the repeating_params_list."""
+        if isinstance(params, list):
+            return hash(tuple([self._hash_params(dictionary) for dictionary in params]))
+        else:
+            return hash(tuple([item for _, item in params.items()]))
+
     @staticmethod
     def _fill_params_dict(template: dict, *args, **kwargs) -> dict:
         BaseDispersion._guard_invalid_params(list(kwargs.keys()), list(template.keys()))
@@ -56,6 +65,8 @@ def _fill_params_dict(template: dict, *args, **kwargs) -> dict:
 
         for i, val in enumerate(args):
             key = list(template.keys())[i]
+            if isinstance(val, Parameter):
+                val = val.value
             params[key] = val
             pos_arguments.add(key)
 
@@ -64,6 +75,8 @@ def _fill_params_dict(template: dict, *args, **kwargs) -> dict:
                 raise InvalidParameters(
                     f"Parameter {key} already set by positional argument"
                 )
+            if isinstance(value, Parameter):
+                value = value.value
             params[key] = value
 
         return params
@@ -80,6 +93,10 @@ def __init__(self, *args, **kwargs):
             if self.single_params[param] is None:
                 raise InvalidParameters(f"Please specify parameter {param}")
 
+        self.last_lbda = None
+        self.hash_single_params = None
+        self.hash_rep_params = None
+
     @abstractmethod
     def dielectric_function(self, lbda: npt.ArrayLike) -> npt.NDArray:
         """Calculates the dielectric function in a given wavelength window.
@@ -114,6 +131,39 @@ def get_dielectric(self, lbda: Optional[npt.ArrayLike] = None) -> npt.NDArray:
         """Returns the dielectric constant for wavelength 'lbda' default unit (nm)
         in the convention ε1 + iε2."""
         lbda = self.default_lbda_range if lbda is None else lbda
+
+        from .table_epsilon import TableEpsilon
+        from .table_index import Table
+
+        if not isinstance(self, (DispersionSum, IndexDispersionSum)):
+            if isinstance(self, (TableEpsilon, Table)):
+                if self.last_lbda is lbda:
+                    return self.cached_diel
+                else:
+                    self.last_lbda = lbda
+                    self.cached_diel = np.asarray(
+                        self.dielectric_function(lbda), dtype=np.complex128
+                    )
+                    return self.cached_diel
+            else:
+                new_single_hash = self._hash_params(self.single_params)
+                new_rep_hash = self._hash_params(self.rep_params)
+
+                if (
+                    self.last_lbda is lbda
+                    and self.hash_single_params == new_single_hash
+                    and self.hash_rep_params == new_rep_hash
+                ):
+                    return self.cached_diel
+                else:
+                    self.last_lbda = lbda
+                    self.hash_single_params = new_single_hash
+                    self.hash_rep_params = new_rep_hash
+                    self.cached_diel = np.asarray(
+                        self.dielectric_function(lbda), dtype=np.complex128
+                    )
+                    return self.cached_diel
+
         return np.asarray(self.dielectric_function(lbda), dtype=np.complex128)
 
     def get_refractive_index(self, lbda: Optional[npt.ArrayLike] = None) -> npt.NDArray:

tests/benchmark_fitting.py

@@ -94,12 +94,12 @@ def test_fitting_structure_updates(benchmark, datadir):
 
     @fit(psi_delta, params)
     def model(lbda, params):
-        SiO2.single_params["n0"] = params["SiO2_n0"]
-        SiO2.single_params["n1"] = params["SiO2_n1"]
-        SiO2.single_params["n2"] = params["SiO2_n2"]
-        SiO2.single_params["k0"] = params["SiO2_k0"]
-        SiO2.single_params["k1"] = params["SiO2_k1"]
-        SiO2.single_params["k2"] = params["SiO2_k2"]
+        SiO2.single_params["n0"] = params["SiO2_n0"].value
+        SiO2.single_params["n1"] = params["SiO2_n1"].value
+        SiO2.single_params["n2"] = params["SiO2_n2"].value
+        SiO2.single_params["k0"] = params["SiO2_k0"].value
+        SiO2.single_params["k1"] = params["SiO2_k1"].value
+        SiO2.single_params["k2"] = params["SiO2_k2"].value
 
         layer.set_thickness(params["SiO2_d"])