JEDI log files ingest class#

The Joint Effort for Data assimilation Integration (JEDI) data assimilation system uses the Object Oriented Prediction System (OOPS) at its core to handle most high level functionality. During the minimization of variational data assimilation, OOPS will write to a log file information that can be parsed and visualized. This class parses the logfile and reads the data into an eva-compatible format for use in transfomrs and plotting.

JEDI log files injest for batch processing#

An example YAML file may look like the following:

datasets:
  - type: JediLog
    collection_name: jedi_log_test_rpcg
    jedi_log_to_parse: ./jedi_log.var_rpcg.txt
    data_to_parse:
      convergence: true

The above keys are defined as follows:

  • collection_name: the name of the dataset that gets propagated throughout eva

  • type: JediLog the class to parse an OOPS log

  • jedi_log_to_parse: path to log file to parse

  • data_to_parse: dictionary of options to parse Currently supported types of data to parse are as follows:

    • convergence: true or false

Following the eva convention, datasets will be available to the transforms and plots sections like so: name::group::variable where:

  • name is the collection_name defined in the YAML

  • group is the type of data that is parsed (under the data_to_parse key)

  • variable is the name of the variable that is parsed (see source code for details)

JEDI log files injest for interactive processing#

To apply the JEDI log files injest class during an eva interactive, call the load_collection() with the JEDI log files name:

    EvaInteractive.load_collection(collection_name, filenames, 'JediLog')

where

  • collection_name is a user-defined label for the loaded collection

  • filenames is a file path string or list of strings to be loaded into the data collection

  • 'JediLog' selects the JEDI log files injest class