Overflow information in dbschema output
The last portion of the dbschema output shows values that have many duplicates.
The number of duplicates of indicated values must be greater than
a critical amount that is determined as approximately 25 percent of
the resolution times the number of rows. If left in the general distribution
data, the duplicates would skew the distribution, so they are moved
from the distribution to a separate list, as the following example
shows:
--- OVERFLOW ---
1: ( 5, 56)
2: ( 6, 63)For this example, the critical amount is 0.25 * 0.10 * 165, or 4.125. Therefore, any value that is duplicated five or more times is listed in the overflow section. Two values in this distribution are duplicated five or more times in the table: the value 56 is duplicated five times, and the value 63 is duplicated six times.