Janos Gabler
Janos Gabler
* estimagic version used, if any: 0.4.6 * Python version, if any: 3.11 * Operating System: Linux ### What would you like to enhance and why? Is it related to...
Currently, some documentation notebooks (e.g. [shapley basic spotify](https://github.com/aai-institute/pyDVL/blob/develop/notebooks/shapley_basic_spotify.ipynb)) do not pass seeds or random number generators when calculating shapley values. We should go over all documentation notebooks and make sure...
The test case `test_shapley_batch_size[1-PermutationSampler-beta_coefficient_w-5-test_game0]` sometimes fails due to a precision problem. The test output is: ```txt x: array([0.171717, 0.23 , 0.272727, 0.646465, 0.686869]) y: array([0.171717, 0.222222, 0.272727, 0.646465, 0.68 ])...
The link to the 'example' initialization file in the Tutorial of the online documentation is broken.
### Description This PR refactors the least-core to make use of the new samplers in `pydvl.valuation`. The previously supported least-core methods are now just special cases: "exact" is obtained using...
Release 0.5.0 will bring several breaking changes and deprecations. On a high level, the major changes are: - Implement EP-02: Static typing - Implement EP-03: Alignment with SciPy - Rename...
## Description of the algorithm The KNN Shapley method is based on this [paper](https://arxiv.org/pdf/1908.08619) The basic algorithm is: The shapley values are calculated independently for each test data point and...
The new design of data valuation methods avoids repeated computations of the utility function without relying on caching. We could therefore get rid of our current caching implementation based on...