Streaming Histograms - Calculating Online Histograms

Question

I am looking for an algorithm to generate a histogram for a large amount of streaming data, max and min are not known in advance, but the standard deviation and average value are in a certain range.

I appreciate your ideas.

Greetings

Answer 1

I found one solution. Sec. 2.2 "A real-time histogram bar from a parallel decision tree algorithm of a parallel solution." Algo is implemented by the NumericHistogram class in the Hive project:

, . : - -, " ", J. Machine Learning 11 (2010), . 849--872. , (, 20-80) .

Answer 2

. , . , , . , ( ) , .

: . , . .

Answer 3

, "GoHistogram", (NumericHistogram Weighted Numeric Histogram). (https://code.google.com). :

https://github.com/VividCortex/gohistogram