Miguel de Benito Delgado
Miguel de Benito Delgado
In the "setup python" github action, both dev and docs requirements [are always installed](https://github.com/aai-institute/pyDVL/blob/f4ebb0eaa46b1d7ade249eb19c0fc310f12cbe9a/.github/actions/python/action.yml#L18). But this action is used e.g. in the build docs stage [in the main workflow](https://github.com/aai-institute/pyDVL/blob/f4ebb0eaa46b1d7ade249eb19c0fc310f12cbe9a/.github/workflows/main.yaml#L60), where...
The new Glossary page makes the Methods page a bit redundant because it contains descriptions of some of the methods as well as links to the implementations. Perhaps we can...
See [the paper](https://arxiv.org/abs/1802.06640) and [reference repo](https://github.com/bsharchilev/influence_boosting) (abandoned, not typed, not packaged => fix in the repo and import as library if we can have ownership / admin rights. However, there's...
For completeness in the implementation of all methods in _Yan, Tom, and Ariel D. Procaccia. “If You Like Shapley Then You’ll Love the Core.” In Proceedings of the 35th AAAI...
We could add factories for `GroupedDataset` which group by categorical feature, or by label, e.g. ```python data = GroupedDataset.by_label(dataset) data2 = GroupedDataset.by_feature(dataset, feature_name) data3 = GroupedDataset.by_feature(dataset, feature_name, bins=10) # For...
Using [nptyping and beartype](https://github.com/ramonhagenaars/nptyping/blob/master/USERDOCS.md#Quickstart) we can pin down the dimensions that we accept. For instance, the constructor of `ValuationResult` can only accept 1D arrays but we don't enforce that. Alternatively,...
We don't always use the default value of the Scorer. For instance when iterating over a permutation, we might want to do: ```python prev_score = u({}) ``` Then we also...
Like in mkdocs-material's docs. See https://github.com/squidfunk/mkdocs-material/discussions/6544
- point removal - mislabeled data detection - ...