def forward(self, x_): # Theano
non_root_proportions, root_height = self.topology.get_proportions(x_)
if self.max_height is None:
root_val = tt.log(root_height)
else:
root_val = logit(
root_height /
(self.max_height - self.topology.get_max_leaf_height()))
return tt.concatenate(
[logit(non_root_proportions),
tt.stack(root_val)])
def __init__(self, p=None, logit_p=None, *args, **kwargs):
super().__init__(*args, **kwargs)
if sum(int(var is None) for var in [p, logit_p]) != 1:
raise ValueError('Specify one of p and logit_p')
if p is not None:
self._is_logit = False
self.p = p = tt.as_tensor_variable(floatX(p))
self._logit_p = logit(p)
else:
self._is_logit = True
self.p = tt.nnet.sigmoid(floatX(logit_p))
self._logit_p = tt.as_tensor_variable(logit_p)
self.mode = tt.cast(tround(self.p), 'int8')
def __init__(self, p=None, logit_p=None, *args, **kwargs):
super().__init__(*args, **kwargs)
if sum(int(var is None) for var in [p, logit_p]) != 1:
raise ValueError('Specify one of p and logit_p')
if p is not None:
self._is_logit = False
self.p = p = tt.as_tensor_variable(p)
self._logit_p = logit(p)
else:
self._is_logit = True
self.p = tt.nnet.sigmoid(logit_p)
self._logit_p = tt.as_tensor_variable(logit_p)
self.mode = tt.cast(tround(self.p), 'int8')
def forward(self, x):
return logit(x)
def forward(self, rv_var, rv_value):
return logit(rv_value)
