Add warning, if fixed parameter is supposed to be profiled (#1155)

* Replaced logger.warn with loggerwarning as warn is depracated.

* Replaced logger.warn with logger.warning as warn is depracated.

* Added warning if a fixed parameter is called for profiling.

---------

Co-authored-by: Daniel Weindl <[email protected]>

pypesto/history/optimizer.py

@@ -134,7 +134,7 @@ def finalize(self, message: str = None, exitflag: int = None):
             and not allclose(result[X], self.x_min)
         ):
             # issue a warning, as if this happens, then something may be wrong
-            logger.warn(
+            logger.warning(
                 f"History has a better point {fval} than the current best "
                 "point {self.fval_min}."
             )

pypesto/optimize/optimizer.py

@@ -608,7 +608,7 @@ def get_default_options(self):
     def check_x0_support(self, x_guesses: np.ndarray = None) -> bool:
         """Check whether optimizer supports x0."""
         if x_guesses is not None and x_guesses.size > 0:
-            logger.warn("The Dlib optimizer does not support x0.")
+            logger.warning("The Dlib optimizer does not support x0.")
         return False
 
 
@@ -666,7 +666,7 @@ def is_least_squares(self):
     def check_x0_support(self, x_guesses: np.ndarray = None) -> bool:
         """Check whether optimizer supports x0."""
         if x_guesses is not None and x_guesses.size > 0:
-            logger.warn("The pyswarm optimizer does not support x0.")
+            logger.warning("The pyswarm optimizer does not support x0.")
         return False
 
 
@@ -940,7 +940,7 @@ def is_least_squares(self):
     def check_x0_support(self, x_guesses: np.ndarray = None) -> bool:
         """Check whether optimizer supports x0."""
         if x_guesses is not None and x_guesses.size > 0:
-            logger.warn("The pyswarms optimizer does not support x0.")
+            logger.warning("The pyswarms optimizer does not support x0.")
         return False
 
 
@@ -1186,7 +1186,7 @@ def check_x0_support(self, x_guesses: np.ndarray = None) -> bool:
             nlopt.GN_DIRECT_L_RAND_NOSCAL,
         ):
             if x_guesses is not None and x_guesses.size > 0:
-                logger.warn(
+                logger.warning(
                     f"The NLopt optimizer method {self.method} does "
                     "not support x0."
                 )

pypesto/profile/profile.py

@@ -135,6 +135,10 @@ def create_next_guess(
     for i_par in profile_index:
         # only compute profiles for free parameters
         if i_par in problem.x_fixed_indices:
+            # log a warning
+            logger.warning(
+                f"Parameter {i_par} is fixed and will not be profiled."
+            )
             continue
 
         current_profile = result.profile_result.get_profiler_result(