IXSEL and TBSEL are the index and table selectivity respectively for an access path to a table. For example, consider accessing a table of SALES via an index on the RETURN_CODE column for the predicate RETURN_CODE = DAMAGED. The index selectivity might be 7.5000e-01, indicating that 75% of the rows in the index have this value. But if only a quarter of a percent of SALES rows have non-null RETURN_CODE values, the TBSEL would be 1.8750e-03.

ORIG CDN (original cardinality) is the number of rows in the table. CMPTD CDN (computed cardinality) is the number of rows from the table that are estimated to satisfy the non-join predicates against that table.

DENS is the density of the column. In a histogram any value that occurs as the end point for more than one bucket is a "popular" value. For popular values, the selectivity of a predicate is estimated to be the proportion of histogram endpoints that it spans. For non-popular values, this density function is used. For example, if a column has 10 popular values and 75% of the rows in table have popular values for this column, and if there are 1000 distinct values for the column, then the density is (1 - 0.75) / (1000 - 10). That is, it is the proportion of non-popular values divided by the number of non-popular values. This is taken to be the selectivity of predicates for non-popular values.

I set the event 10053 event for a session and I could not understand a few terms in the resulting trace file
a) How is IXSEL different from TBSEL.
b) What do the ORIG CDN, CMPTD CDN and DENS stand for ?

Please answer only if you have time to spare.