ENIR .fit returns an error when AUC equals 1.

[eugeneberta]

Hi, thanks for the great package!

It seems netcal.binning.ENIR returns an error when .fit is called with separable binary probabilities (predictions for which AUC = 1).
Here is a minimal example to reproduce the bug:

import numpy as np
from netcal.binning import ENIR
cal = ENIR()
cal.fit(np.array([0,1]), np.array([0,1]))

returns the error message

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
Cell In[1], [line 4](vscode-notebook-cell:?execution_count=1&line=4)
      2 from netcal.binning import ENIR
      3 cal = ENIR()
----> [4](vscode-notebook-cell:?execution_count=1&line=4) cal.fit(np.array([0,1]), np.array([0,1]))

File /opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/Decorator.py:62, in dimensions.<locals>.check_dim.<locals>.new_f(*args, **kwds)
     60     elif type(d) == int:
     61         assert len(a.shape) == d, "dimension of arg \'%s\' must match %s but is %d" % (f.__code__.co_varnames[i], str(d), len(a.shape))
---> [62](https://file+.vscode-resource.vscode-cdn.net/opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/Decorator.py:62) return f(*args, **kwds)

File /opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/binning/ENIR.py:270, in ENIR.fit(self, X, y)
    267 model_list.pop()
    269 # get model scores and binning models by elbow method
--> [270](https://file+.vscode-resource.vscode-cdn.net/opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/binning/ENIR.py:270) self._model_scores, self._binning_models = self._elbow(X, y, model_list, self.score_function, alpha=0.001)
    271 return self

File /opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/Decorator.py:35, in accepts.<locals>.check_accepts.<locals>.new_f(*args, **kwds)
     33     else:
     34         assert isinstance(a, t), "arg \'%s\' does not match %s" % (f.__code__.co_varnames[i],t)
---> [35](https://file+.vscode-resource.vscode-cdn.net/opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/Decorator.py:35) return f(*args, **kwds)

File /opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/AbstractCalibration.py:499, in AbstractCalibration._elbow(self, confidences, ground_truth, model_list, score_function, alpha)
    473 @accepts(np.ndarray, np.ndarray, list, str, float)
    474 def _elbow(self, confidences: np.ndarray, ground_truth: np.ndarray, model_list: List['AbstractCalibration'],
    475            score_function: str = 'BIC', alpha: float = 0.001) -> Tuple[List[float], List['AbstractCalibration']]:
    476     """
    477     Select models by Bayesian score and discard models below a certain threshold with elbow method.
    478 
   (...)
    496         Tuple with two lists of (normed) model scores and list of models.
    497     """
--> [499](https://file+.vscode-resource.vscode-cdn.net/opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/AbstractCalibration.py:499)     model_scores = self._calc_model_scores(confidences, ground_truth, model_list, score_function)
    501     num_models = len(model_scores)
    502     score_variance = np.var(model_scores)

File /opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/Decorator.py:35, in accepts.<locals>.check_accepts.<locals>.new_f(*args, **kwds)
     33     else:
     34         assert isinstance(a, t), "arg \'%s\' does not match %s" % (f.__code__.co_varnames[i],t)
---> [35](https://file+.vscode-resource.vscode-cdn.net/opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/Decorator.py:35) return f(*args, **kwds)

File /opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/AbstractCalibration.py:469, in AbstractCalibration._calc_model_scores(self, confidences, ground_truth, model_list, score_function)
    466     raise ValueError("Unknown score function \'%s\'. Fix your implementation")
    468 # calculate relative likelihood of each model and normalize scores
--> [469](https://file+.vscode-resource.vscode-cdn.net/opt/anaconda3/envs/calib/lib/python3.12/site-packages/netcal/AbstractCalibration.py:469) model_scores = np.exp((np.min(score) - score) / 2.)
    471 return model_scores

File /opt/anaconda3/envs/calib/lib/python3.12/site-packages/numpy/core/fromnumeric.py:2953, in min(a, axis, out, keepdims, initial, where)
   2836 @array_function_dispatch(_min_dispatcher)
   2837 def min(a, axis=None, out=None, keepdims=np._NoValue, initial=np._NoValue,
   2838         where=np._NoValue):
   2839     """
   2840     Return the minimum of an array or minimum along an axis.
   2841 
   (...)
   2951     6
   2952     """
-> [2953](https://file+.vscode-resource.vscode-cdn.net/opt/anaconda3/envs/calib/lib/python3.12/site-packages/numpy/core/fromnumeric.py:2953)     return _wrapreduction(a, np.minimum, 'min', axis, None, out,
   2954                           keepdims=keepdims, initial=initial, where=where)

File /opt/anaconda3/envs/calib/lib/python3.12/site-packages/numpy/core/fromnumeric.py:88, in _wrapreduction(obj, ufunc, method, axis, dtype, out, **kwargs)
     85         else:
     86             return reduction(axis=axis, out=out, **passkwargs)
---> [88](https://file+.vscode-resource.vscode-cdn.net/opt/anaconda3/envs/calib/lib/python3.12/site-packages/numpy/core/fromnumeric.py:88) return ufunc.reduce(obj, axis, dtype, out, **passkwargs)

ValueError: zero-size array to reduction operation minimum which has no identity