-
Notifications
You must be signed in to change notification settings - Fork 174
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Cannot supply array of eval_points to EnsembleArray #766
Labels
Comments
tbekolay
added a commit
that referenced
this issue
Mar 29, 2017
This is essentially a no-op distribution that takes a set of samples and provides them when the `samples` method is called. It's intended to simplify situations in which a distribution or samples could be provided by converting samples to a distribution. The main place this happens is when passing `Ensemble` arguments to the `EnsembleArray` constructor. Previously, a distribution had to be provided, since the passed parameters are set as defaults on the `EnsembleArray` network. Now, any samples passed in are wrapped in the `Samples` distribution, which makes it possible to pass in samples. Tests have been added to verify this. Addresses #691 and #766.
adityagilra
pushed a commit
to adityagilra/nengo
that referenced
this issue
Jun 21, 2017
This is essentially a no-op distribution that takes a set of samples and provides them when the `samples` method is called. It's intended to simplify situations in which a distribution or samples could be provided by converting samples to a distribution. The main place this happens is when passing `Ensemble` arguments to the `EnsembleArray` constructor. Previously, a distribution had to be provided, since the passed parameters are set as defaults on the `EnsembleArray` network. Now, any samples passed in are wrapped in the `Samples` distribution, which makes it possible to pass in samples. Tests have been added to verify this. Addresses nengo#691 and nengo#766.
Has this been resolved with the merge of #1233? |
ping @arvoelke |
Yep!
Thanks. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Possible solutions?
Force people to only use distributions with EnsembleArrays. This seems like an unreasonable limitation to me.
Include the number of dimensions before the ensembles are configured, i.e.
But this is hacky because this solution has to be repeated for all similar situations with other networks. Also, it's important to note that you can't just do
ens_kwargs['dimensions'] = ens_dimensions
otherwise you get undefined behaviour, because it relies on the order of dictionary iteration to updatedimensions
beforeeval_points
, which may or may not happen depending on the alignment of the sun with Neptune.Create a new
StaticDistribution
that always iterates through a fixed list, and then casteval_points
arrays to this distribution when the param is set (thus removingDistOrArrayParam
)? I think this is a better long-term solution, assuming I'm not overlooking anything.Note: Add a unit test for this.
The text was updated successfully, but these errors were encountered: