Consider the following 2-D ARRAY data type and table.
CREATE TYPE shot_ary AS VARRAY(1:50)(1:50) OF INTEGER; CREATE TABLE seismic_table ( id INTEGER, shots shot_ary);
The table is populated with the following values:
/* The first 2 elements are populated; the rest are uninitialized. */ INSERT INTO seismic_table VALUES (1, shot_ary(11, 12)); /* Empty ARRAY instance */ INSERT INTO seismic_table VALUES (2, shot_ary()); /* Update the empty ARRAY instance such that element  is set to a value; then elements  and  are set to NULL, the rest are uninitialized */ UPDATE seismic_table SET shots = 1133 WHERE id = 2;
The following query returns the number of elements in the shots array that contain NULL or values that are non-NULL.
SELECT id, shots.OCOUNT() FROM seismic_table;
The following is the result of the query.
ID shots.OCOUNT() -- -------------- 1 2 (the first 2 elements contain values that are non-NULL) 2 3 (the first 3 elements contain NULL or values that are non-NULL)
The following is the same query using function-style syntax.
SELECT id, OCOUNT(shots) FROM seismic_table;