Skip to content

Instantly share code, notes, and snippets.

Embed
What would you like to do?
import theano
import theano.tensor as tt
import numpy as np
import pymc3 as pm
print(theano.__version__)
x = np.asarray([0,1])
print(x.dtype)
import numpy as np
test_values = np.arange(32752, 32772, 1);
for v in test_values:
model = pm.Model();
with model:
mu = pm.Normal("mu", mu=0, sd=10);
Y = pm.NegativeBinomial("Y", mu=2**mu, alpha=10, observed=v)
Z = pm.Poisson('Z', mu=2**mu, observed=v)
for RV in model.basic_RVs:
if RV.name in ["Y","Z"]:
print(v, RV.logp(model.test_point));
print(tt.as_tensor_variable(test_values).dtype)
factln = pm.distributions.dist_math.factln
n = 32768 + 10
k = 32768
for val in [factln(n), factln(k), factln(n - k)]:
print(val.eval())
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment