Custom datasets

Description

I followed the description to create a custom dataset here This one does not seem to be up to date.

Context

So that others avoids this problem by following the above example.

Possible Implementation

I needed to run the init of the inherrited class AbstractVersionedDataSet to get this working:

def __init__(self, filepath: str, version: Version = None):
        """Creates a new instance of SimulationDataSet to load / save simulation results (hff5-files) from Seaman at the given filepath.

        Args:
            filepath: The location of the simulation file to load / save data.
        """

        super().__init__(
            filepath=PurePosixPath(filepath),
            version=version,
        )

Possible Alternatives

I or someone else could perhaps update the example similar to the suggestion above?

[martinlarsalbert]

Aha then I guess that it works, if you say so, thanks! By the way if I want to install “latest” instead of “stable” can I do that from pypi or should I get it from GitHub instead?

[AntonyMilneQB]

@martinlarsalbert the difference is that in the latest documentation the dataset inherits from AbstractDataSet rather than AbstractVersionedDataSet. AbstractDataSet doesn’t define an __init__, so no need to do super().__init__.

Don’t worry about the commit, it’s not related - the notification is just an artefact of the way we sync between different repositories.