US 12,130,792 B2
Data merging in distributed computing system
Xing Wei, Xi'an (CN); Xiao Bin Sun, Xi'an (CN); Zhe Shao, Xi'an (CN); Dong Hai Yu, Xi'an (CN); Liu Zhen Duo, Xi'an (CN); and Chun Lei Xu, Xi'an (CN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Dec. 2, 2021, as Appl. No. 17/457,301.
Prior Publication US 2023/0177035 A1, Jun. 8, 2023
Int. Cl. G06F 16/22 (2019.01); G06F 16/27 (2019.01)
CPC G06F 16/2272 (2019.01) [G06F 16/27 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A computer implemented method for managing datasets for a histogram, the computer implemented method comprising a number of processing units in a computer:
determining a first span for first bins containing first datapoints in a first dataset in the datasets, wherein the first span is determined based on a first distribution of the first datapoints in the first dataset and a desired number of bins;
adding a datapoint to the first datapoints in the first dataset, wherein the first distribution of the first datapoints has a lower bound and an upper bound comprising the first datapoints;
adjusting the lower bound and the upper bound for the first distribution of the first datapoints with the datapoint to form an adjusted distribution for the first datapoints;
determining a second adjusted span for the first dataset based on the adjusted distribution and the desired number of bins;
adjusting the first bins based on the adjusted distribution and the second adjusted span to form an adjusted first bins containing the first datapoints;
adjusting a second span for second bins containing second datapoints in a second dataset in the datasets to form an adjusted span that matches the first span for the first bins, wherein the second span is determined based on a second distribution of the second datapoints in the second dataset, and wherein the second distribution differs from the first distribution; and
merging the first datapoints in the first bins having the first span with the second datapoints in the second bins having the adjusted span to form a merged dataset for the histogram.