Adaptive Server versions 15.0.2 and later allow selectivity estimates to have the same accuracy as the binary character set, without requiring an excessive number of histogram steps. This benefits queries like the following, which uses range predicates:
select * from t1 where charcolumn > "LMC0021" and charcolumn <= "LMC0029"
If ranges specified falls into the same histogram cell, Adaptive Server can much more accurately estimate this selectivity.
In versions of Adaptive Server earlier than 15.0.2, only the default binary character set benefited from histogram interpolation, which is used to estimate the selectivity of range predicates. For all other character sets, Adaptive Server made a selectivity estimate of 50 percent for a histogram cell. This typically required Adaptive Server to use a large number of histogram cells for character column histograms to reduce the error associated with this estimate.
