Using CPnest to calculate Bayes Theorem evidence, receiving AttributeError: 'MetropolisHastingsSampler' object has no attribute 'thread_id'

Question

I am currently trying to use CPnest to calculate some evidences for given a data set (Bayesian statistics) Upon my first running of the code, I received the error

RuntimeError: 
            Attempt to start a new process before the current process
            has finished its bootstrapping phase.
            This probably means that you are on Windows and you have
            forgotten to use the proper idiom in the main module:
                if __name__ == '__main__':
                    freeze_support()
                    ...
            The "freeze_support()" line can be omitted if the program
            is not going to be frozen to produce a Windows executable.

I looked up the soltuion to this and tried implementing it into my code and no longer got the error, here is my code I am trying to use (the original code that got the error is just the code below without the if __name__ == '__main__': :

import cpnest
from cpnest.model import Model



class ModelClass(Model):
    def __init__(self, times=time1, counts = data10): #time1 and data10 are just simple arrays
        self.counts = counts
        self.times = times
        self.names = ['l0']
        self.bounds=[[0,10]]
            
    def log_likelihood(self,params):
        model = model1(self.counts, params['l0'])
        return np.sum(np.log(model)) - 5*np.log(2*np.pi)
    
if __name__ == '__main__':
    my_model=ModelClass()
    nest = cpnest.CPNest(my_model)
    nest.run()

However, when running this code, I receive a new error for each process-process?

Traceback (most recent call last):
  File "C:\Users\Rp199\anaconda3\lib\multiprocessing\process.py", line 315, in _bootstrap
    self.run()
  File "C:\Users\Rp199\anaconda3\lib\multiprocessing\process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "C:\Users\Rp199\anaconda3\lib\site-packages\cpnest\sampler.py", line 173, in produce_sample
    self._produce_sample()
  File "C:\Users\Rp199\anaconda3\lib\site-packages\cpnest\sampler.py", line 185, in _produce_sample
    self.reset()
  File "C:\Users\Rp199\anaconda3\lib\site-packages\cpnest\sampler.py", line 113, in reset
    for n in tqdm(range(self.poolsize), desc='SMPLR {} init draw'.format(self.thread_id),
AttributeError: 'MetropolisHastingsSampler' object has no attribute 'thread_id'

I am not sure on how to interpret this error and I don't know if it's because I used if __name__ == '__main__': incorrectly or some other issue that I can't see.

Thanks

score 1 · Answer 1 · answered Jan 24 '21 at 17:10

I think this could be a bug with cpnest or one of its dependencies. What Python version are you using and what is your operating system?

Try running the code in Python 3.7 or below. I seem to get the same error in Python 3.8 but not in 3.7 (on MacOS). It seems related to the Python version and the multiprocessing module. It's probably worth raising this issue on their GitHub (https://github.com/johnveitch/cpnest/issues) if you have time, especially if it is inconvenient for you to run the code in a different Python version.

You were right to use if __name__ == "__main__": the way you did, this is typically good practice because it allows you to import the classes and function you defined in the file into other files. When you run the file directly (e.g. $ python my_file.py in the terminal) then the variable __name__ of the space is called '__main__'. When you import my_file elsewhere, the variable __name__ is equal to 'my_file' and thus the condition is false.

I had some buddies that used MacOS and seemed to work fine so I do think it's an issue with multiprocessing module. When I next use cpnest I'll try run it on Python 3.7 and if the issue does persist, i will raise up the issue on Github, thanks! — Rikesh Patel, Jan 29 '21 at 12:05
No worries, I hope it helped and interesting that it's not an OS thing. It's probably just a compatibility issue with Python 3.8 — AlexL, Feb 01 '21 at 10:19

Using CPnest to calculate Bayes Theorem evidence, receiving AttributeError: 'MetropolisHastingsSampler' object has no attribute 'thread_id'

1 Answers1