Closed takashi-ishio closed 3 years ago
I have confirmed that some events are recorded at least 3,309,630,164 times (that cannot be represented by signed 32-bit int)
I confirmed that the revised version recorded 8,473,986,052 events for a data ID using Math 8b.
k-shimari reported that a test
org/apache/commons/math3/optimization/direct/BOBYQAOptimizerTest
in Defects4J 2.0 Math8b might execute some instructions more than 2G times. The recendtata.txt file is like this: