Python-Pymc3-贝叶斯模型

Question

我正在按照本教程使用贝叶斯统计信息来辅助A / B测试。 https://medium.com/@thibalbo/coding-bayesian-ab-tests-in-python-e89356b3f4bd我遇到以下错误

 ---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
~\Anaconda3\pymc3\model.py in get_context(cls)
    179         try:
--> 180             return cls.get_contexts()[-1]
    181         except IndexError:

IndexError: list index out of range

During handling of the above exception, another exception occurred:
TypeError                                 Traceback (most recent call last)
<ipython-input-14-55619ff3a91c> in <module>()
      1 # define metrics
----> 2 pm.Deterministic('difference', prior_v2 - prior_v1)
      3 pm.Deterministic('relation', (prior_v2/prior_v1) - 1)

~\Anaconda3\pymc3\model.py in Deterministic(name, var, model)
   1415     var : var, with name attribute
   1416     """
-> 1417     model = modelcontext(model)
   1418     var.name = model.name_for(name)
   1419     model.deterministics.append(var)

~\Anaconda3\pymc3\model.py in modelcontext(model)
    188     """
    189     if model is None:
--> 190         return Model.get_context()
    191     return model
    192 

~\Anaconda3\pymc3\model.py in get_context(cls)
    180             return cls.get_contexts()[-1]
    181         except IndexError:
--> 182             raise TypeError("No context on context stack")
    183 
    184 

TypeError: No context on context stack

这是我的代码，在“定义指标”之后引发错误：

%matplotlib inline
import pymc3 as pm
import seaborn as sb

n = 1000
A = 680
B = 700

with pm.Model() as model: # context management
    # define priors
    prior_v1 = pm.Beta('prior_v1', alpha=2, beta=2)
    prior_v2 = pm.Beta('prior_v2', alpha=2, beta=2)

    # define likelihood
    like_v1 = pm.Binomial('like_v1', n=n, p=prior_v1, observed=A)
    like_v2 = pm.Binomial('like_v2', n=n, p=prior_v2, observed=B)
# define metrics
pm.Deterministic('difference', prior_v2 - prior_v1)
pm.Deterministic('relation', (prior_v2/prior_v1) - 1)