Collecting statistics on geospatial columns can improve database performance with geospatial data by providing demographic data that can aid the Teradata Database Optimizer in generating the most efficient access and join plans. The basic syntax used for the COLLECT STATISTICS statement applied to geospatial data columns is the same as that used for non-geospatial data, with the following restrictions:
- The COLLECT STATISTICS USING options are not supported on geospatial columns.
- Automated statistics collection and management is not supported for geospatial columns.
- Because geospatial indexes are always NUSIs, the UNIQUE keyword cannot be used.
- The ALL and PARTITION options are not supported for geospatial data.
- The ORDER BY VALUES and ORDER BY HASH clauses are not supported for geospatial data.
- Statistics cannot be collected on geospatial columns in global temporary or volatile tables. Consequently, the TEMPORARY keyword is not supported for geospatial data.
- Statistics collection for geospatial data is supported only for single columns in user tables. It is not supported on geospatial data in views.
- Statistics cannot be collected on geospatial data in join or hash indexes.
- INITIATE INDEX ANALYSIS and RESTART INDEX ANALYSIS are not supported for geospatial indexes.
- If statistics are collected on 3D geospatial data, only the x and y coordinate data are considered. The z coordinate data is ignored. The output of the SHOW STATISTICS command for 3D geospatial data is unchanged from that of 2D geospatial data.