add unc-related type hints

becquerel/core/fitting.py

@@ -967,7 +967,7 @@ def _likelihood_residual(self, params, data, weights, **kwargs):
     def eval(self, x, params=None, **kwargs):
         return self.model.eval(x=x, params=params, **kwargs)
 
-    def calc_area_and_unc(self, component=None, x=None):
+    def calc_area_and_unc(self, component=None, x=None) -> ufloat:
         """Calculate the area (and uncertainty) under the fit (or component thereof).
 
         Parameters
@@ -1056,7 +1056,7 @@ def _calc_area(param_vec, **kwargs):
         # the binwidth, we double the counts per bin but halve the number of bins.
         return ufloat(area, np.sqrt(area_variance))
 
-    def param_val(self, param):
+    def param_val(self, param) -> float:
         """
         Value of fit parameter `param`
         """
@@ -1075,7 +1075,7 @@ def param_val(self, param):
             raise FittingError(f"Unknown param: {param}")
         raise FittingError(f"Unknown backend: {self.backend}")
 
-    def param_unc(self, param):
+    def param_unc(self, param) -> float:
         """
         Fit error of fit parameter `param`
         """
@@ -1095,7 +1095,7 @@ def param_unc(self, param):
             raise FittingError(f"Unknown param: {param}")
         raise FittingError(f"Unknown backend: {self.backend}")
 
-    def param_rel_unc(self, param):
+    def param_rel_unc(self, param) -> float:
         """
         Relative error of fit parameter `param`
         """
@@ -1110,7 +1110,7 @@ def param_val_and_unc(self, param: str) -> ufloat:
         return None
 
     @property
-    def best_values(self):
+    def best_values(self) -> dict[str, float]:
         """Wrapper for dictionary of best_values."""
         if "lmfit" in self.backend:
             return self.result.best_values
@@ -1119,7 +1119,7 @@ def best_values(self):
         raise FittingError(f"Unknown backend: {self.backend}")
 
     @property
-    def init_values(self):
+    def init_values(self) -> dict[str, float]:
         """Wrapper for dictionary of init_values."""
         if "lmfit" in self.backend:
             return self.result.init_values
@@ -1128,7 +1128,7 @@ def init_values(self):
         raise FittingError(f"Unknown backend: {self.backend}")
 
     @property
-    def success(self):
+    def success(self) -> bool:
         if "lmfit" in self.backend:
             return self.result.success
         elif "minuit" in self.backend:
@@ -1144,7 +1144,7 @@ def covariance(self):
             return self.result.covariance
         raise FittingError(f"Unknown backend: {self.backend}")
 
-    def param_dataframe(self, sort_by_model=False):
+    def param_dataframe(self, sort_by_model: bool = False) -> pd.DataFrame:
         """
         Dataframe of all fit parameters value and fit error
         """
@@ -1166,7 +1166,7 @@ def param_dataframe(self, sort_by_model=False):
             )
         return df
 
-    def compute_residuals(self, residual_type="abs"):
+    def compute_residuals(self, residual_type: str = "abs") -> np.ndarray:
         """Compute residuals between the data and the fit.
 
         Parameters