Adding a data product to the standard
A data product can be added to the standard by supplying a yaml formatted file describing the product in a format understood by the vocal tool, implementing all required metadata in this standard. The definition should take the following format:
meta:
file_pattern: "some_pattern_{date}_{flight_number}.nc"
long_name: "Some data product"
short_name: "some_data_product"
description: "A longer description of the data product"
references:
- title: "Some reference"
web: "http://some.url"
doi: "10.1234/5678"
attributes:
Conventions: "CF-1.6 ACDD-1.3"
acknowledgements: "Some acknowledgement"
# ... other attributes
dimensions:
- name: time
size: null # null (or blank) means unlimited
- name: height
size: 100
variables:
- meta:
name: "some_variable"
datatype: "<datatype>" # e.g. float32, int32, str, etc.
dimensions: # Subset of dimensions defined above
- time
- height
attributes:
long_name: "Some variable"
units: "m/s"
# ... other attributes
# ... other variables
groups: # optional
# nested groups, following the same structure as the root group above
Attributes can be specified as literals, or it may be indicated that they are going to be generated on-the-fly during data processing. In this case, the attribute should be specified using a derived_from_file indicator. For example:
attributes:
comment: "<str: derived_from_file>" # A string defined at runtime
revision: "<int8: derived_from_file>" # An integer defined at runtime
revision_comment: "<str: derived_from_file optional>" # An optional string defined at runtime"
actual_range: "<Array[float32]: derived_from_file>" # An array of floats defined at runtime