Am Dienstag, 6. Mai 2008 schrieb Martijn van Oosterhout:
> Cross-table correlations are easy for the second part, because it's
> fairly simple to see where it could be used. However, no-one has come
> up with an algorithm to produce a useful number to use. For others it's
> harder.
For an algorithm, principal components analysis would be my guess. It is
designed to answer the question "this column value is tied to this other
column value in this way" [quote Simon], at least for the sort of data that a
B-tree would cover. For nonlinear data, it is of course harder.
> In general postgres could use many bits of information not currently
> available. For example: A=B implies lower(A)=lower(B), hence an index
> on lower(A) could be used to optimise comparisons against A. Certain
> operations preserve order, which may also be useful.
Locale horrors looming ... ;-)
