Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

No generation progress and ABC.Sampler INFO: Could not discard from active set

See original GitHub issue

Hello! I run several workers in parallel and although enough particles are accepted per generation the next generation does not start automatically. In order to start the next generation I have to manually reset the workers. I, too, get the following error for individual workers frequently and was wondering whether these issues might be connected:

ABC.Sampler INFO: Could not discard from active set: Doesn't exist.
Process Process-5:
Traceback (most recent call last):
  File "/home/bq_anetzl/.conda/envs/pyABC_ABcDModel/lib/python3.9/multiprocessing/process.py", line 315, in _bootstrap
    self.run()
  File "/home/bq_anetzl/.conda/envs/pyABC_ABcDModel/lib/python3.9/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "/home/bq_anetzl/.conda/envs/pyABC_ABcDModel/lib/python3.9/site-packages/pyabc/sampler/redis_eps/cli.py", line 137, in _work
    work_on_population_dynamic(
  File "/home/bq_anetzl/.conda/envs/pyABC_ABcDModel/lib/python3.9/site-packages/pyabc/sampler/redis_eps/work.py", line 83, in work_on_population_dynamic
    while get_int(N_ACC) < get_int(N_REQ) and (
  File "/home/bq_anetzl/.conda/envs/pyABC_ABcDModel/lib/python3.9/site-packages/pyabc/sampler/redis_eps/work.py", line 36, in get_int
    return int(redis.get(idfy(var, ana_id, t)).decode())
AttributeError: 'NoneType' object has no attribute 'decode'

I hope you can help me figure out how to fix my problem. Thank you!

Issue Analytics

State:
Created 2 years ago
Comments:5 (3 by maintainers)

Top GitHub Comments

1reaction

yannikschaeltecommented, Aug 3, 2021

Hi, good to hear it works. 10 acceptances would for usual applications be far too few (100-10000 are more common), maybe manually run you model a few times and note down average run times. pyABC has been used for anything between milliseconds and 3h, however for anything beyond a few seconds runtime, we usually use massive parallelization with a few hundred cores. If your manually calculated run time does not fit the 20h, there may still be sth wrong, e.g. with the priors, leading to many rejected particles.

0reactions

anetzlcommented, Aug 3, 2021

Hi, I updated pyABC and ran it as recommended - it works! It takes a very long time for all workers to finish (20 h from gen 0 to gen 1 with only 10 Acceptances required) but I guess that is due to my model. Resetting the remaining workers does not give above error message. Thanks a lot!

Top Results From Across the Web

ABC Officer form - New York State Liquor Authority

Completed Certification of Appointment of Alcoholic Beverage Control Officer signed by both the proposed ABC Officer and a current Officer of ...

System Center vNext Data Protection Manager error codes

Lists error codes that may occur in System Center vNext Data Protection Manager.

Component Reference - Apache JMeter - User's Manual

Several test elements use JMeter properties to control their behaviour. These properties are normally resolved when the class is loaded.

Customer File Custom Audiences - Marketing API

To improve how advertisers create and manage their audiences, Customer File Custom Audiences that have not been used in any active ad sets...

ABC-(6421, 6422, 6423)-SDS.pdf - Fiberlock

Hazards Not Otherwise Classified (NHOC): Not applicable. Other Information: Repeated or prolonged skin contact may cause allergic reactions with susceptible.