timescaledb_information.dimensions

Get metadata about the dimensions of hypertables, returning one row of metadata for each dimension of a hypertable. For a time-and-space-partitioned hypertable, for example, two rows of metadata are returned for the hypertable.

A time-based dimension column has either an integer datatype (bigint, integer, smallint) or a time-related datatype (timestamptz, timestamp, date). The time_interval column is defined for hypertables that use time datatypes. Alternatively, for hypertables that use integer datatypes, the integer_interval and integer_now_func columns are defined.

For space-based dimensions, metadata is returned that specifies their number of num_partitions. The time_interval and integer_interval columns are not applicable for space based dimensions.

Available columns

NameTypeDescription
hypertable_schemaTEXTSchema name of the hypertable
hypertable_nameTEXTTable name of the hypertable
dimension_numberBIGINTDimension number of the hypertable, starting from 1
column_nameTEXTName of the column used to create this dimension
column_typeREGTYPEType of the column used to create this dimension
dimension_typeTEXTIs this time based or space based dimension?
time_intervalINTERVALTime interval for primary dimension if the column type is based on Postgres time datatypes
integer_intervalBIGINTInteger interval for primary dimension if the column type is an integer datatype
integer_now_funcTEXTinteger_now function for primary dimension if the column type is integer based datatype
num_partitionsSMALLINTNumber of partitions for the dimension

Sample usage

Get information about the dimensions of hypertables.

  2. --Create a time and space partitioned hypertable
  4. CREATE TABLE dist_table(time timestamptz, device int, temp float);
  6. SELECT create_hypertable('dist_table', 'time',  'device', chunk_time_interval=> INTERVAL '7 days', number_partitions=>3);
  8. SELECT * from timescaledb_information.dimensions
  10.   ORDER BY hypertable_name, dimension_number;
  12. -[ RECORD 1 ]-----+-------------------------
  14. hypertable_schema | public
  16. hypertable_name   | dist_table
  18. dimension_number  | 1
  20. column_name       | time
  22. column_type       | timestamp with time zone
  24. dimension_type    | Time
  26. time_interval     | 7 days
  28. integer_interval  |
  30. integer_now_func  |
  32. num_partitions    |
  34. -[ RECORD 2 ]-----+-------------------------
  36. hypertable_schema | public
  38. hypertable_name   | dist_table
  40. dimension_number  | 2
  42. column_name       | device
  44. column_type       | integer
  46. dimension_type    | Space
  48. time_interval     |
  50. integer_interval  |
  52. integer_now_func  |
  54. num_partitions    | 2

Get information about dimensions of a hypertable that has two time-based dimensions.

  2. CREATE TABLE hyper_2dim (a_col date, b_col timestamp, c_col integer);
  4. SELECT table_name from create_hypertable('hyper_2dim', 'a_col');
  6. SELECT add_dimension('hyper_2dim', 'b_col', chunk_time_interval=> '7 days');
  8. SELECT * FROM timescaledb_information.dimensions WHERE hypertable_name = 'hyper_2dim';
  10. -[ RECORD 1 ]-----+----------------------------
  12. hypertable_schema | public
  14. hypertable_name   | hyper_2dim
  16. dimension_number  | 1
  18. column_name       | a_col
  20. column_type       | date
  22. dimension_type    | Time
  24. time_interval     | 7 days
  26. integer_interval  |
  28. integer_now_func  |
  30. num_partitions    |
  32. -[ RECORD 2 ]-----+----------------------------
  34. hypertable_schema | public
  36. hypertable_name   | hyper_2dim
  38. dimension_number  | 2
  40. column_name       | b_col
  42. column_type       | timestamp without time zone
  44. dimension_type    | Time
  46. time_interval     | 7 days
  48. integer_interval  |
  50. integer_now_func  |
  52. num_partitions    |