SKOptLearner doesn't work for multi variate domain

[caenrigen]

The code below shows the issue.
I will do a pull request for a possible fix

import adaptive
from adaptive.learner import SKOptLearner
def F_1D(x):
    return x**2
def F_2D(x):
    return x[0]**2 + x[1]**2

# Works
learner = SKOptLearner(F_1D, dimensions=[(-20., 20.)])
runner = adaptive.runner.simple(learner, goal=lambda l: l.npoints > 10)
# Doesn't work
learner = SKOptLearner(F_2D, dimensions=[(-20., 20.),(-10., 10.)])
runner = adaptive.runner.simple(learner, goal=lambda l: l.npoints > 10)

Error output:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-5-1fecee9ed826> in <module>
     35 try:
     36     learner = SKOptLearner(F_2D, dimensions=[(-20., 20.),(-10., 10.)])
---> 37     runner = adaptive.runner.simple(learner, goal=lambda l: l.npoints > 10)
        global runner = None
        global adaptive.runner.simple = <function simple at 0xb1565f8c0>
        global learner = <adaptive.learner.skopt_learner.SKOptLearner object at 0x1a197df990>
        global goal = undefined
        global l = undefined
        global l.npoints = undefined
     38 except:
     39     text_file = open("exeption.txt", "w")

/usr/local/anaconda3/envs/conda-dicarlo-lab/lib/python3.7/site-packages/adaptive/runner.py in simple(learner=<adaptive.learner.skopt_learner.SKOptLearner object>, goal=<function <lambda>>)
    681         for x in xs:
    682             y = learner.function(x)
--> 683             learner.tell(x, y)
        learner.tell = <bound method SKOptLearner.tell of <adaptive.learner.skopt_learner.SKOptLearner object at 0x1a197df990>>
        x = [-3.866859822710609, 2.529043603793383]
        y = 21.34866643838174
    684 
    685 

/usr/local/anaconda3/envs/conda-dicarlo-lab/lib/python3.7/site-packages/adaptive/learner/skopt_learner.py in tell(self=<adaptive.learner.skopt_learner.SKOptLearner object>, x=[-3.866859822710609, 2.529043603793383], y=21.34866643838174, fit=True)
     30 
     31     def tell(self, x, y, fit=True):
---> 32         self.pending_points.discard(x)
        self.pending_points.discard = <built-in method discard of set object at 0x1c1c69e500>
        x = [-3.866859822710609, 2.529043603793383]
     33         super().tell([x], y, fit)
     34 

TypeError: unhashable type: 'list'

[jbweston]

Closed by #234