How to use the contextualbandits.utils._FixedPredictor function in contextualbandits

def _apply_softmax(x):
    x[:, :] = np.exp(x - x.max(axis=1).reshape((-1, 1)))
    x[:, :] = x / x.sum(axis=1).reshape((-1, 1))
    return None

class _FixedPredictor:
    def __init__(self):
        pass

    def fit(self, X=None, y=None, sample_weight=None):
        pass

    def decision_function_w_sigmoid(self, X):
        return self.decision_function(X)

class _BetaPredictor(_FixedPredictor):
    def __init__(self, a, b):
        self.a = a
        self.b = b

    def predict_proba(self, X):
        preds = np.random.beta(self.a, self.b, size = X.shape[0]).reshape((-1, 1))
        return np.c_[1.0 - preds, preds]

    def decision_function(self, X):
        return np.random.beta(self.a, self.b, size = X.shape[0])

    def predict(self, X):
        return (np.random.beta(self.a, self.b, size = X.shape[0])).astype('uint8')

    def predict_avg(self, X):
        pred = self.decision_function(X)

def predict_avg(self, X):
        pred = self.decision_function(X)
        _apply_inverse_sigmoid(pred)
        return pred

    def predict_rnd(self, X):
        return self.predict_avg(X)

    def predict_ucb(self, X):
        return self.predict_avg(X)

    def exploit(self, X):
        return np.repeat(self.a / self.b, X.shape[0])

class _ZeroPredictor(_FixedPredictor):

    def predict_proba(self, X):
        return np.c_[np.ones((X.shape[0], 1)),  np.zeros((X.shape[0], 1))]

    def decision_function(self, X):
        return np.zeros(X.shape[0])

    def predict(self, X):
        return np.zeros(X.shape[0])

    def predict_avg(self, X):
        return np.repeat(-1e6, X.shape[0])

    def predict_rnd(self, X):
        return self.predict_avg(X)

def decision_function(self, X):
        return np.zeros(X.shape[0])

    def predict(self, X):
        return np.zeros(X.shape[0])

    def predict_avg(self, X):
        return np.repeat(-1e6, X.shape[0])

    def predict_rnd(self, X):
        return self.predict_avg(X)

    def predict_ucb(self, X):
        return self.predict_avg(X)

class _OnePredictor(_FixedPredictor):

    def predict_proba(self, X):
        return np.c_[np.zeros((X.shape[0], 1)),  np.ones((X.shape[0], 1))]

    def decision_function(self, X):
        return np.ones(X.shape[0])

    def predict(self, X):
        return np.ones(X.shape[0])

    def predict_avg(self, X):
        return np.repeat(1e6, X.shape[0])

    def predict_rnd(self, X):
        return self.predict_avg(X)

def decision_function(self, X):
        return np.ones(X.shape[0])

    def predict(self, X):
        return np.ones(X.shape[0])

    def predict_avg(self, X):
        return np.repeat(1e6, X.shape[0])

    def predict_rnd(self, X):
        return self.predict_avg(X)

    def predict_ucb(self, X):
        return self.predict_avg(X)

class _RandomPredictor(_FixedPredictor):
    def _gen_random(self, X):
        return np.random.random(size = X.shape[0])

    def predict(self, X):
        return (self._gen_random(X) >= .5).astype('uint8')

    def decision_function(self, X):
        return np.random.random(size = X.shape[0])

    def predict_proba(self, X):
        pred = self._gen_random(X)
        return np.c[pred, 1 - pred]

class _BootstrappedClassifierBase:
    def __init__(self, base, nsamples, percentile = 80, partialfit = False, partial_method = "gamma", njobs = 1):
        self.bs_algos = [deepcopy(base) for n in range(nsamples)]