ImportMLModel

Source

from CommonServerPython import *

# import json

# Python template - reading arguments, calling a command, handling errors and returning results


def read_file_content(input_entry_or_string):
    res = demisto.getFilePath(input_entry_or_string)
    if not res:
        return_error(f"Entry {input_entry_or_string} not found")
    file_path = res["path"]
    with open(file_path) as f:
        file_content = f.read()
    return file_content


def main():
    entry_id = demisto.args()["entryID"]
    model_name = demisto.args()["modelName"]
    storing_method = demisto.args()["modelStoreType"]
    encoded_file_content = read_file_content(entry_id)
    file_content = json.loads(encoded_file_content)
    args = {
        "modelData": file_content["modelData"],
        "modelName": model_name,
        "modelLabels": file_content["model"]["labels"],
        "modelOverride": True,
    }
    if storing_method == "mlModel":
        res = demisto.executeCommand("createMLModel", args)
        if is_error(res):
            return_error(get_error(res))
        confusion_matrix = file_content["model"]["evaluation"]["confusionMatrix"]
        res = demisto.executeCommand(
            "evaluateMLModel",
            {
                "modelConfusionMatrix": confusion_matrix,  # disable-secrets-detection
                "modelName": model_name,
            },
        )
    elif storing_method == "list":
        res = demisto.executeCommand("createList", {"listName": model_name, "listData": file_content["modelData"]})
    else:
        return_error(
            f'Unsupported *modelStoreType* value received ({storing_method}). *modelStoreType* should be "mlModel" or "list"'
        )

    if is_error(res):
        return_error(get_error(res))
    demisto.results("done")


if __name__ in ["__main__", "__builtin__", "builtins"]:
    main()

README