improver.wxcode.wxcode_decision_tree module

Module containing weather symbol decision tree

improver.wxcode.wxcode_decision_tree.wxcode_decision_tree()

Define queries that comprise the weather symbol decision tree. Each queries contains the following elements:

• succeed: The next query to call if the diagnostic being queried

  satisfies the current query.

• fail: The next query to call if the diagnostic being queried

  does not satisfy the current query.

• diagnostic_missing_action: For optional diagnostic data

  What to do if the diagnostic is missing. It can take the keywords succeed or fail. The logic will then follow that path if the diagnostic is missing.

• probability_thresholds: A list of probability thresholds that the

  query requires. One entry is provided for each diagnostic field being tested.

• threshold_condition: The condition the diagnostic must satisfy

  relative to the probability threshold (e.g. greater than (>) the probability threshold).

• condition_combination: The way (AND, OR) in which multiple

  conditions should be combined; e.g. rainfall > 0.5 AND snowfall > 0.5

• diagnostics_fields: The diagnostics which are being used in the

  query. If this is a list of lists, the two fields in a given list are subtracted (1st - (2nd * gamma)) and then compared with the probability threshold.

• diagnostic_gamma (NOT UNIVERSAL): This is the gamma factor that

  is used when comparing two fields directly, rather than comparing a single field to a probability threshold. e.g. gamma * P(SnowfallRate) < P(RainfallRate).

• diagnostic_thresholds: The thresholding that is expected to have

  been applied to the input data; this is used to extract the appropriate data from the input cubes.

• diagnostic_conditions: The condition that is expected to have

  been applied to the input data; this can be used to ensure the thresholding is as expected.

Returns

A dictionary containing the queries that comprise the decision tree.

Return type

dict