aws/random-cut-forest-by-aws

Expose Summarize as a generic

sudiptoguha opened this issue · 1 comments

The summarization function is useful in the missing value interpolation/forecast for multiple variables since it accounts for scenarios. It Is an useful summarization primitive because it does not need to know the target number of clusters/scenarios and may be used elsewhere. The summarization already accepts arbitrary distance functions. Goal is to generalize to instead of being fixed to <float[]>.

PR 347