Good evening!

I am attempting to use the example found at http://docs.pymc.io/notebooks/dependent_density_regression.html# to model a simpler data set so I know what is going on within the example. I am able to successfully run the example above itself so it must be an issue on my end.

```
import pymc3 as pm
import numpy as np
import matplotlib.pyplot as plt
from theano import shared
import theano.tensor as tt
import pandas as pd
x1 = np.linspace(0., 9.9, 10)
x2 = np.linspace(10., 30.9, 30)
x = np.concatenate((x1, x2), axis = 0)
y1 = 10.*x1 + 5 + np.random.ranf(x1.size)*100.
y2 = y1[-1:] + 40.*x2 + -40.*x2[0] + np.random.ranf(x2.size)*300.
y = np.concatenate((y1, y2), axis = 0)
plt.plot(x1, y1, linestyle = 'None', marker = '.')
plt.plot(x2, y2, linestyle = 'None', marker = '.')
def norm_cdf(z):
return 0.5 * (1 + tt.erf(z / np.sqrt(2)))
def stick_breaking(v):
return v * tt.concatenate([tt.ones_like(v[:, :1]), tt.extra_ops.cumprod(1. - v, axis=1)[:, :-1]], axis=1)
x = x[:, np.newaxis]
y = y[:, np.newaxis]
X = shared(x, broadcastable=(False, True))
K = 20
with pm.Model() as model:
alpha = pm.Normal('alpha', 0., 1., shape=K)
beta = pm.Normal('beta', 0., 1., shape=K)
v = norm_cdf(alpha + beta * X)
w = pm.Deterministic('w', stick_breaking(v))
gamma = pm.Normal('gamma', 0., 10., shape=K)
delta = pm.Normal('delta', 0., 10., shape=K)
mu = pm.Deterministic('mu', gamma + delta * X)
tau = pm.Gamma('tau', 1., 1., shape=K)
obs = pm.NormalMixture('obs', w, mu, tau=tau, observed=y)
with model:
trace = pm.sample(20000, init='advi+adapt_diag', progressbar = True)
```

The error is: “Auto-assigning NUTS sampler…

Initializing NUTS using advi+adapt_diag…

ERROR (theano.gof.opt): Optimization failure due to: local_grad_log_erfc_neg

ERROR (theano.gof.opt): node: Elemwise{true_div,no_inplace}(Elemwise{mul,no_inplace}.0, Elemwise{erfc,no_inplace}.0)

ERROR (theano.gof.opt): TRACEBACK:

ERROR (theano.gof.opt): Traceback (most recent call last):

File “/Users/dhooghe/anaconda3/lib/python3.6/site-packages/theano/gof/opt.py”, line 2019, in process_node

replacements = lopt.transform(node)

File “/Users/dhooghe/anaconda3/lib/python3.6/site-packages/theano/tensor/opt.py”, line 6780, in local_grad_log_erfc_neg

if not exp.owner.inputs[0].owner:

AttributeError: ‘NoneType’ object has no attribute ‘owner’”

I am fundamentally missing something here so any illumination that you could give would be greatly appreciated. Thank you!