Hiebl opened a new pull request #1517: URL: https://github.com/apache/systemds/pull/1517
This patch extends the built-in functions transformencode and transformapply by equi-height binning on the local runtime. For this purpose the selected column gets sorted, then the bin boundaries are calulated as quantiles which then are converted to inout indicies and applied as bin boundaries analogous to already existing equi-width binning. The bining problem for non divisible bin numbers has been adressed with spillover. This has the advantage that all but the last bin a gurantued to be equi-height. The drawback is that the last bin has equi-height + inputlength % bin number height. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
