NEWS for anomalyDetection package
Changes for version 0.2.2
- Better tolerance in
mahalanobis_distance when inverting covariance matrices.
horns_curve have been rewritten in C++ using the
RcppArmadillo package. This greatly improved the speed (and accuracy) of these functions.
tabulate_state_vector has been rewritten using the
dplyr package, greatly improving the speed of this function. Greater traceability is now also present for missing values and numeric variables.
- Producing histogram matrix to visually display anomalous blocks made easier with addition of
- Properly registered native routines and disabled symbol search.