def _():
v0 = qy.select(True, 3, 4)
v1 = qy.select(False, 3, 4)
@qy.python(v0, v1)
def _(v0_py, v1_py):
result[0] = v0_py
result[1] = v1_py
def ll(self, parameter, sample, out):
"""
Compute constant-distribution log-likelihood.
"""
qy.select(
parameter.data.load() == sample.data.load(),
0.0,
numpy.finfo(float).min,
) \
.store(out)
def log_add_d(x_in, y_in):
s = x_in >= y_in
a = qy.select(s, x_in, y_in)
@qy.if_else(a == -numpy.inf)
def _(then):
if then:
qy.return_(-numpy.inf)
else:
qy.return_(a + qy.log1p(qy.exp(qy.select(s, y_in, x_in) - a)))
def log_add_d(x_in, y_in):
s = x_in >= y_in
a = qy.select(s, x_in, y_in)
@qy.if_else(a == -numpy.inf)
def _(then):
if then:
qy.return_(-numpy.inf)
else:
qy.return_(a +
qy.log1p(qy.exp(qy.select(s, y_in, x_in) - a)))
def _(then):
if then:
qy.return_(-numpy.inf)
else:
qy.return_(a +
qy.log1p(qy.exp(qy.select(s, y_in, x_in) - a)))
def __abs__(self):
"""
Return the absolute value of this value.
"""
return qy.select(self > 0.0, self, -self)
def _():
qy.return_(qy.select(k == n, 0.0, -numpy.inf))
def _(then):
if then:
qy.return_(-numpy.inf)
else:
qy.return_(a + qy.log1p(qy.exp(qy.select(s, y_in, x_in) - a)))
def _():
qy.return_(qy.select(k == n, 0.0, -numpy.inf))
