An implementation of the Random Cut Forest data structure for sketching streaming data, with support for anomaly detection, density estimation, imputation, and more.
This shows up in PR 317 -- due to the implicit mass calculation (to save space) and the serialization. In version 2.0, the mass at leaves was explicitly maintained but that information was already available in the corresponding sampler. This was remedied, but not addressed in serialization of single node forests.
This shows up in PR 317 -- due to the implicit mass calculation (to save space) and the serialization. In version 2.0, the mass at leaves was explicitly maintained but that information was already available in the corresponding sampler. This was remedied, but not addressed in serialization of single node forests.