Helper Functions

top-level utility functions for common operations such as key generation, epoch parsing and local model prediction.

Functions

create a WizataDSAPIClient from environment variables. This is the recommended way to create a client.

return: WizataDSAPIClient instance.

import wizata_dsapi

client = wizata_dsapi.api()
client.info()

generate a unique key for experiment, pipeline, model, template or other entities. The key is a human-readable combination of date, color and animal.

return: unique key string.

import wizata_dsapi

key = wizata_dsapi.generate_unique_key()
# e.g. '0218_red_penguin_042'

generate an epoch timestamp in milliseconds based on a relative datetime string (e.g. now+6h, now-7d).

Name	Type	Default	Description
formatted_string	str		formatted epoch representation using relative time (e.g. 'now-7d', 'now+6h').
now		None	override the current datetime (defaults to `datetime.now(timezone.utc)`).
return: epoch in milliseconds.

Supported units: y = 365d, M = 30d, w = 7d, d = 24h, h = 60m, m = 60s, s = 1000ms.

import wizata_dsapi

epoch_ms = wizata_dsapi.generate_epoch("now-7d")

verify if a string matches the supported relative datetime format (e.g. now-7d, now+6h).

Name	Type	Default	Description
formatted_string	str		string to verify.
return: True if format is valid, False otherwise.

execute a machine learning model locally on a dataframe.

Name	Type	Default	Description
df	pandas.DataFrame		dataframe to use as input.
model_info	ModelInfo		model information handler (must have trained_model loaded).
mapping_table		None	optional mapping table for column renaming.
return: output dataframe with predicted values.

import wizata_dsapi

model = wizata_dsapi.api().download_model("my_model")
result = wizata_dsapi.predict(df, model)