Jargon buster #

C #

config.yaml

a file that specifies the metadata used during file lookup (for read, write, and register), used to configure the overall API behaviour.

D #

Data product

objects that are accessable with the standard API read_xxx(), write_xxx() calls (i.e. toml or hdf5 files)

Data registry

a database for recording metadata about data and model outputs

E #

External object

an object that is not in an internal pipeline format (i.e. not toml or hdf5 files)

F #

finalise()

calculates the hash of the external objects / data products and renames files, check files exist, etc.

I #

initialise()

reads in the config.yaml file, generates a new entry in code_run and returns id, creates new submission_script (and if necessary creates a new code_repo) entry in local registry

L #

link_read()

connects code_run to external_object as input and returns path location (see also link_write())

link_write()

connects code_run to data_product as output and returns path location (see also link_read())

R #

read:

a field in config.yaml used to specify the reading of an object (a data product or an external object) that belongs to the user (see also write: and register:)

register:

a field in config.yaml used to specify the registration of an object (usually an external object) that exists elsewhere (see also write: and read:)

add_to_register()

registers the object(s) listed in register: (in the config.yaml file)

W #

write:

a field in config.yaml used to specify the writing of an object (a data product or an external object) that belongs to the user (see also register: and read:)