AttributeError: 'StanModel' object has no attribute 'fit_class'

[allisonwang]

I kept getting this error when trying to fit prophet model.

AttributeError                            Traceback (most recent call last)
<ipython-input-139-a8e399fdfab2> in <module>()
----> 1 m.fit(df)

~/env/lib/python3.6/site-packages/fbprophet/forecaster.py in fit(self, df, **kwargs)
   1001             try:
   1002                 params = model.optimizing(
-> 1003                     dat, init=stan_init, iter=1e4, **kwargs)
   1004             except RuntimeError:
   1005                 params = model.optimizing(

~/env/lib/python3.6/site-packages/pystan/model.py in optimizing(self, data, seed, init, sample_file, algorithm, verbose, as_vector, **kwargs)
    510             data = {}
    511         seed = pystan.misc._check_seed(seed)
--> 512         fit = self.fit_class(data, seed)
    513 
    514         m_pars = fit._get_param_names()

AttributeError: 'StanModel' object has no attribute 'fit_class'

Has anyone encountered this issue before? What could be causing this error?

[bletham]

Usually this is because PyStan is not working correctly. Could you run this code to check if it is?

import pystan
model_code = 'parameters {real y;} model {y ~ normal(0,1);}'
model = pystan.StanModel(model_code=model_code)  # this will take a minute
y = model.sampling(n_jobs=1).extract()['y']
y.mean()  # should be close to 0

Most commonly the issue with PyStan is getting the C++ compiler set up (especially in windows). For that I would refer to (https://pystan.readthedocs.io/en/latest/windows.html) . If you're using Anaconda, then it is best to install fbprophet through conda forge, as described in https://facebook.github.io/prophet/docs/installation.html

[allisonwang]

thanks! I am installing this in Linux with pip but there might be something wrong when building the package. I will look into this.