Cannot resolve inputs error with Stochastic Volatility model

Questions
1

Hi!

I tried to implement the “Stochastic Volatility model” from the examples section of the website.
Copy & paste doesn’t work because the jupyter kernel crashes at this code block

with stochastic_vol_model:
prior = pm.sample_prior_predictive(500)

with the exception:

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
~/anaconda3/envs/datascience/lib/python3.7/site-packages/pymc3/distributions/distribution.py in _draw_value(param, point, givens, size)
575                 try:
--> 576                     return dist_tmp.random(point=point, size=size)
577                 except (ValueError, TypeError):

~/anaconda3/envs/datascience/lib/python3.7/site-packages/pymc3/distributions/continuous.py in random(self, point, size)
   1960         nu, mu, lam = draw_values([self.nu, self.mu, self.lam],
-> 1961                                   point=point, size=size)
   1962         return generate_samples(stats.t.rvs, nu, loc=mu, scale=lam**-0.5,

~/anaconda3/envs/datascience/lib/python3.7/site-packages/pymc3/distributions/distribution.py in draw_values(params, point, size)
412             if to_eval == missing_inputs:
--> 413                 raise ValueError('Cannot resolve inputs for {}'.format([str(params[j]) for j in to_eval]))
414             to_eval = set(missing_inputs)

ValueError: Cannot resolve inputs for ['Elemwise{mul,no_inplace}.0']

During handling of the above exception, another exception occurred:

ValueError                                Traceback (most recent call last)
<ipython-input-234-9d6a8559045e> in <module>
  1 with btc_model:
----> 2     prior = pm.sample_prior_predictive(50)

~/anaconda3/envs/datascience/lib/python3.7/site-packages/pymc3/sampling.py in sample_prior_predictive(samples, model, vars, var_names, random_seed)
   1320     names = get_default_varnames(model.named_vars, include_transformed=False)
   1321     # draw_values fails with auto-transformed variables. transform them later!
-> 1322     values = draw_values([model[name] for name in names], size=samples)
   1323 
   1324     data = {k: v for k, v in zip(names, values)}

~/anaconda3/envs/datascience/lib/python3.7/site-packages/pymc3/distributions/distribution.py in draw_values(params, point, size)
393                                         point=point,
394                                         givens=temp_givens,
--> 395                                         size=size)
396                     givens[next_.name] = (next_, value)
397                     drawn[(next_, size)] = value

~/anaconda3/envs/datascience/lib/python3.7/site-packages/pymc3/distributions/distribution.py in _draw_value(param, point, givens, size)
583                     with _DrawValuesContextBlocker():
584                         val = np.atleast_1d(dist_tmp.random(point=point,
--> 585                                                             size=None))
586                     # Sometimes point may change the size of val but not the
587                     # distribution's shape

~/anaconda3/envs/datascience/lib/python3.7/site-packages/pymc3/distributions/continuous.py in random(self, point, size)
   1959         """
   1960         nu, mu, lam = draw_values([self.nu, self.mu, self.lam],
-> 1961                                   point=point, size=size)
   1962         return generate_samples(stats.t.rvs, nu, loc=mu, scale=lam**-0.5,
   1963                                 dist_shape=self.shape,

~/anaconda3/envs/datascience/lib/python3.7/site-packages/pymc3/distributions/distribution.py in draw_values(params, point, size)
411         while to_eval or missing_inputs:
412             if to_eval == missing_inputs:
--> 413                 raise ValueError('Cannot resolve inputs for {}'.format([str(params[j]) for j in to_eval]))
414             to_eval = set(missing_inputs)
415             missing_inputs = set()

ValueError: Cannot resolve inputs for ['Elemwise{mul,no_inplace}.0']

I’m using Python 3.6 with pymc3 3.7.

Any ideas? Found a bug report where someone suggested that changing variable string names…but not for me unfortunately :frowning:

Thanks!

2

Could you try upgrading pymc3 to master or the newest release? This issue is already fixed i think.