You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In many examples and tests map_reduce is passed a utility, but this works only when num_jobs <= num_runs. When num_jobs > num_runs, data being a Utility is not supported. Considering we probably want to keep data a simple collection, find a way around calling map_reduce with a utility and test it.
In order not to copy all the data to all the workers, map reduce was splitting the dataset in many different chunks. This is a problem, since each worker was calculating the Shapley values on a different dataset, causing wrong results.
We have now switched to ray which uses Plasma under to hood to share objects with the various workers.
In many examples and tests map_reduce is passed a utility, but this works only when num_jobs <= num_runs. When num_jobs > num_runs, data being a Utility is not supported. Considering we probably want to keep data a simple collection, find a way around calling map_reduce with a utility and test it.
See, e.g. this discussion
Once fixed, unify interface in montecarlo methods (there is a TODO in permutation_montecarlo_shapley)
The text was updated successfully, but these errors were encountered: