Introduce 'tell_pending' which replaces 'tell(x, None)'

and I also tackle

• lower_bound <= learner.ask()[0] <= upper_bound
• if we tell(x, y) with x outside the bounds, the point should be added to data, but not any of the other internal datastructures. This will mean that the point will be "invisible" to the learning algorithm, and only useful for plotting and data extraction

for both the Learner1D and Learner2D in that MR. I included this in here to avoid merge conflicts later.

adaptive/learner/skopt_learner.py

@@ -27,11 +27,13 @@ class SKOptLearner(Optimizer, BaseLearner):
         self.function = function
         super().__init__(**kwargs)
 
-    def tell(self, x, y, fit=True):
-        if y is not None:
-            # 'skopt.Optimizer' takes care of points we
-            # have not got results for.
-            super().tell([x], y, fit)
+    def tell(self, x, y, fit=True):    
+        super().tell([x], y, fit)
+
+    def tell_pending(self, x):
+        # 'skopt.Optimizer' takes care of points we
+        # have not got results for.
+        pass
 
     def remove_unfinished(self):
         pass
@@ -46,7 +48,10 @@ class SKOptLearner(Optimizer, BaseLearner):
             # estimator of loss, but it is the cheapest.
             return 1 - model.score(self.Xi, self.yi)
 
-    def ask(self, n, add_data=True):
+    def ask(self, n, tell_pending=True):
+        if not tell_pending:
+            raise NotImplementedError('Asking points is an irreversible '
+                'action, so use `ask(n, tell_pending=True`.')
         points = super().ask(n)
         # TODO: Choose a better estimate for the loss improvement.
         if self.space.n_dims > 1: