Table Blueprint

The table blueprint protocol provides a convention for expressing tabular data in Conduit. Each data entry in a table represents a column and each column contains the same number of rows. Nodes that conform to the table blueprint protocol are easily translated to and from CSV files.


To conform to the table blueprint protocol, a Node must have a “values” child which is a list OR an object and:

  • All of “values” children are data arrays OR mcarrays
  • All of “values” children must have the same number of elements

A node will also conform to the table blueprint protocol if it is a collection of tables. A valid collection of tables must be a list OR an object and:

  • All of its children are valid tables as defined above.

Table Examples

An example of a table blueprint in yaml format:

    scalar_column: [0, 1, 2, 3]
        x: [0, 1, 2, 3]
        y: [0, 1, 2, 3]
        z: [0, 1, 2, 3]

An example of a collection of tables in yaml format:

            x: [0, 1, 2, 3]
            y: [0, 1, 2, 3]
            z: [0, 1, 2, 3]
        scalar_data: [0, 1, 2, 3]
        scalar_data: [0, 1]
            a: [0, 1]
            b: [0, 1]
            c: [0, 1]

The table blueprint namespace includes a function basic(), that generates a simple example of tabular data.

conduit::blueprint::table::examples::basic(conduit::index_t nx,
                                           conduit::index_t ny,
                                           conduit::index_t nz,
                                           Node &res);

This function will generate points (points/x, points/y, points/z) in a uniform manner based off the arguments nx, ny, and nz. Also included in the output table is a point_data column that starts at 0 and increases by 1 for each point.

The resulting data is placed the Node res, which is passed in via a reference.

For more details, see the unit tests that exercise these examples in src/tests/blueprint/t_blueprint_table_verify.cpp and src/tests/blueprint/t_blueprint_table_examples.cpp.