Using batch processing

As mentioned at the end of the previous tutoral, the command line tools produce YAML files that allow it to easily reproduce a project. This is the purpose of the batch processing command yaw_cli run.

All we need to do is run this command on the YAML configuration file (see contents below) found in the previously created project output:

$ yaw_cli run reproduced -s output/setup.yaml

The results will be written to a new project folder called reproduced and will contain the same files in the end, including a copy of the original YAML configuration. More details on configuration files can be found in the section on batch processing.

Contents of output/setup.yaml.

# yet_another_wizz setup configuration (auto generated)

configuration:
    binning:
        zmin: 0.07
        zmax: 1.42
    scales:
        rmin: 100.0
        rmax: 1000.0

data:
    n_patches: 32
    reference:
        data:
            filepath: reference.fits
            ra: ra
            dec: dec
            redshift: z
        rand:
            filepath: randoms.fits
            ra: ra
            dec: dec
            redshift: z
    unknown:
        data:
            filepath: unknown.fits
            ra: ra
            dec: dec

tasks:
  - cross
  - auto_ref
  - zcc
  - plot

Processing subsets

As in the previous example, we can also process the unknown sample in subsets if these subsets are stored in separate input catalogues. In that case, the syntax of the configuration file from above changes slightly since we must assign an index to each of the unknown data input catalogues. These bin indices can be assinged arbitrarily, but to be consistent with the yaw_cli cross example the configuration file would have to look like this:

Contents of output/setup.yaml with tomographic bins.

# yet_another_wizz setup configuration (auto generated)

configuration:
    binning:
        zmin: 0.07
        zmax: 1.42
    scales:
        rmin: 100.0
        rmax: 1000.0

data:
    n_patches: 32
    reference:
        data:
            filepath: reference.fits
            ra: ra
            dec: dec
            redshift: z
        rand:
            filepath: randoms.fits
            ra: ra
            dec: dec
            redshift: z
    unknown:
        data:
            filepath:
                1: unknown1.fits
                2: unknown2.fits
                3: unknown3.fits
            ra: ra
            dec: dec

tasks:
  - cross
  - auto_ref
  - zcc
  - plot

Note

To provide a multiple unknown data subsets the single file path must be replaced by a mapping of bin index to file path, the rest of the file is unchanged compared to the example above.