Repository Issue Activity (beta)

#2001 MinMaxScaler.transform_one produces NaN for unseen features

#1944 Strict typing all modules

#1997 Normalizer.transform_one fails on zero vectors

#1709 Controlling the cleanup of DBStream

#1914 Drift module roadmap: broaden detector coverage + build an evaluation harness (6-month focus)

#1423 Quantile regression

#1992 check_roc_auc passes for a detector that returns a constant score

#1389 Theta forecasting model

#1964 BayesianLinearRegression.predict_one(..., with_dist=True) reports sqrt(sigma) instead of sigma

#1439 Online GapEncoder

#1946 Calibrated forecast features beat StandardScaler pipelines on river's own datasets — evidence, a companion package, and a PR offer

#1919 Migrate all mini-batch (_many) methods to narwhals for dataframe-agnostic support

#1430 Professional-grade mypy conf

#1840 FSRS

#1881 Support dataframe agnostic mini-batching

#1912 Explore: move linear_model mini-batch learning into Rust (true per-sample SGD)

#1576 `TFIDF.transform_many()` fails on `DataFrame` input

#1429 Use stubs files from dependencies

#1435 Create ABI3 (backwards compatible) wheels

#1917 Implement Bayesian logistic regression

#1600 Applying TimeRolling in LogisticRegression

#1472 DBSTREAM: add "time" argument to the learn function

#1395 Chi-squared test

#1428 Implement dataframe interchange API

#1398 Display transformer union elements vertically rather than horizontally