def discrete_likelihood(node, data=None, dtype=np.float64):
probs, marg_ids, observations = leaf_marginalized_likelihood(node, data, dtype)
scipy_obj, params = get_scipy_obj_params(node)
probs[~marg_ids] = scipy_obj.pmf(observations, **params)
probs[probs == 1.0] = 0.999999999
probs[probs == 0.0] = 0.000000001
return probs
def categorical_dictionary_likelihood(node, data=None, dtype=np.float64):
probs, marg_ids, observations = leaf_marginalized_likelihood(
node, data, dtype)
dict_probs = [node.p.get(val, 0.0) for val in observations]
probs[~marg_ids] = dict_probs
return probs
def piecewise_log_likelihood(node, data=None, dtype=np.float64, **kwargs):
probs, marg_ids, observations = leaf_marginalized_likelihood(
node, data, dtype, log_space=True)
probs[~marg_ids] = piecewise_complete_cases_log_likelihood(node,
observations,
dtype=dtype)
return probs
def gamma_likelihood(node, data=None, dtype=np.float64):
probs, marg_ids, observations = leaf_marginalized_likelihood(node, data, dtype)
observations[observations == 0] += POS_EPS
scipy_obj, params = get_scipy_obj_params(node)
probs[~marg_ids] = scipy_obj.pdf(observations, **params)
return probs
def discrete_log_likelihood(node, data=None, dtype=np.float64, **kwargs):
probs, marg_ids, observations = leaf_marginalized_likelihood(
node, data, dtype, log_space=True)
scipy_obj, params = get_scipy_obj_params(node)
probs[~marg_ids] = scipy_obj.logpmf(observations, **params)
# probs[probs == 1.0] = 0.999999999
# probs[np.isinf(probs)] = 0.000000001
probs[np.isinf(probs)] = MIN_NEG # 0.000000001
return probs
def categorical_likelihood(node, data=None, dtype=np.float64):
probs, marg_ids, observations = leaf_marginalized_likelihood(node, data, dtype)
cat_data = observations.astype(np.int64)
assert np.all(np.equal(np.mod(cat_data, 1), 0))
out_domain_ids = cat_data >= node.k
idx_out = ~marg_ids
idx_out[idx_out] = out_domain_ids
probs[idx_out] = 0
idx_in = ~marg_ids
idx_in[idx_in] = ~out_domain_ids
probs[idx_in] = np.array(node.p)[cat_data[~out_domain_ids]]
return probs
def gaussian_likelihood(node,
data=None,
dtype=np.float64,
bmarg=None,
ibm=None):
probs, marg_ids, observations = leaf_marginalized_likelihood(
node, data, dtype)
scipy_obj, params = get_scipy_obj_params(node)
# probs[~marg_ids] = scipy_obj.pdf(observations, **params)
if bmarg:
ibm = ibm[:, node.scope]
probs_reliable = np.expand_dims(scipy_obj.pdf(observations, **params),
axis=1)
probs_unreliable = np.expand_dims(scipy.stats.norm.cdf(
observations, loc=params['loc'], scale=params['scale']),
axis=1)
probs = np.where(ibm, probs_reliable, probs_unreliable)
else:
probs[~marg_ids] = scipy_obj.pdf(observations, **params)
return probs
def continuous_likelihood(node, data=None, dtype=np.float64, **kwargs):
probs, marg_ids, observations = leaf_marginalized_likelihood(
node, data, dtype)
scipy_obj, params = get_scipy_obj_params(node)
probs[~marg_ids] = scipy_obj.pdf(observations, **params)
return probs
def uniform_log_likelihood(node, data=None, dtype=np.float64, **kwargs):
probs, marg_ids, observations = leaf_marginalized_likelihood(
node, data, dtype, log_space=True)
probs[~marg_ids] = np.log(node.density)
return probs
def uniform_likelihood(node, data=None, dtype=np.float64):
probs, marg_ids, observations = leaf_marginalized_likelihood(
node, data, dtype)
probs[~marg_ids] = node.density
return probs
def poisson_likelihood(node, data=None, dtype=np.float64):
probs, marg_ids, observations = leaf_marginalized_likelihood(
node, data, dtype)
scipy_obj, params = get_scipy_obj_params(node)
probs[~marg_ids] = scipy_obj.pmf(observations, **params)
return probs
