diff --git a/ProcessOptimizer/learning/gbrt.py b/ProcessOptimizer/learning/gbrt.py
index 24b3c10f..e2418e0d 100644
--- a/ProcessOptimizer/learning/gbrt.py
+++ b/ProcessOptimizer/learning/gbrt.py
@@ -110,7 +110,7 @@ def predict(self, X, return_std=False, return_quantiles=False):
 
         elif return_std:
             std_quantiles = [0.16, 0.5, 0.84]
-            is_present_mask = np.isin(std_quantiles, self.quantiles)
+            is_present_mask = np.in1d(std_quantiles, self.quantiles)
             if not np.all(is_present_mask):
                 raise ValueError(
                     "return_std works only if the quantiles during "
diff --git a/ProcessOptimizer/optimizer/optimizer.py b/ProcessOptimizer/optimizer/optimizer.py
index fc0231a3..5217e186 100644
--- a/ProcessOptimizer/optimizer/optimizer.py
+++ b/ProcessOptimizer/optimizer/optimizer.py
@@ -431,7 +431,7 @@ def ask(self, n_points=None, strategy="stbr_fill"):
         if not ((isinstance(n_points, int) and n_points > 0) or n_points is None):
             raise ValueError("n_points should be int > 0, got " + str(n_points))
         # These are the only filling strategies which are supported
-
+        
         if strategy == "stbr_full" and self._n_initial_points < 1:
             # Steienerberger sampling can not be used from an empty Xi set
             if self.Xi == []:
@@ -446,13 +446,13 @@ def ask(self, n_points=None, strategy="stbr_fill"):
                 # Returns 'n_points' Steinerberger points
                 X = self.stbr_scipy(n_points=n_points)
             return X
-
+        
         if n_points is None or n_points == 1:
             return self._ask()
-
+        
         # The following assertions deal with cases in which the user asks for more than
         # single experiments
-
+        
         supported_strategies = [
             "cl_min",
             "cl_mean",
@@ -469,7 +469,7 @@ def ask(self, n_points=None, strategy="stbr_fill"):
                 + ", "
                 + "got %s" % strategy
             )
-
+        
         if strategy in ["stbr_fill", "stbr_full"] and self.get_constraints() is not None:
             raise ValueError(
                 "Steinerberger (default setting) sampling can not be used with constraints,\
@@ -599,9 +599,7 @@ def _ask(self):
 
             if abs(min_delta_x) <= 1e-8:
                 warnings.warn(
-                    "FYI: The optimizer already has information about the "
-                    "objective at this point. This can e.g., occur when the "
-                    "optimizer assesses noise and uncertainty."
+                    "The objective has been evaluated " "at this point before."
                 )
 
             # return point computed from last call to tell()
@@ -1017,19 +1015,19 @@ def set_constraints(self, constraints):
         * `constraints` [list] or [Constraints]:
             Can either be a list of Constraint objects or a Constraints object
         """
-
+                
         if self.n_objectives > 1:
             raise RuntimeError(
             "Can't set constraints for multiobjective optimization. The NSGA-II algorithm \
             used for multiobjective optimization does not support constraints."
             )
-
+            
         if self._n_initial_points > 0 and self._lhs:
             raise RuntimeError(
                 "Can't set constraints while latin hypercube sampling points are not exhausted.\
                 Consider reinitialising the optimizer with lhs=False as argument."
             )
-
+            
         if constraints:
             if isinstance(constraints, Constraints):
                 # If constraints is a Constraints object we simply add it