Chris White
Chris White
### Summary This PR is the first in a series of changes that update deployments in a few different ways based on feedback; the current PR: - removes the `manifest.json`...
We should allow for optionally white-listing private methods on a per-class basis for the API reference documentation. This will require updating the format of `docs/outline.toml` to include a list of...
This PR wires up the new `PREFECT_EXPERIMENTAL_ENABLE_NEW_ENGINE` setting to also use the new engine for deployment-triggered flow runs.
@TomAugspurger has mentioned having a first release soon; I'd like to collect the things that we think are blocking that from happening at this moment and check them off the...
We should probably implement abstract base classes for `Regularizer` and `Family` with corresponding documentation; as #47 and #46 have highlighted, the lack of a solid template can lead to problems....
One of the biggest hanging issues is to organize / discuss what the **API** will look like for `dask_glm` and what sorts of things we expose. This decision impacts algorithm...
We need to determine a robust testing framework for our logistic regression algorithms, starting with a review of the accuracy of the current ADMM code. **Unregularized Problems** - For unregularized...
An all-too-often ignored side of optimization is the initialization; there is a lot of research out there suggesting that for both convex (and even more so for non-convex) optimization problems,...
*What is the best way to handle intercepts?* Right now, the algorithms assume the user creates a column of `1`s in their dask array, à la `statsmodels`. However, sometimes it's...
We should have a testing framework / experimentation engine to test the algorithms for robustness to various types of input data (normalized / not-normalized / bad condition numbers / partial...