Hi All, I am facing an issue while executing Zingg in Databricks Runtime 15.4 LTS (Spark 3.5.0) with Zingg Version 0.4.0. Issue Description: I am trying to execute Zingg using the following code: options = ClientOptions([ClientOptions.PHASE, "findTrainingData"]) # Zingg execution for the given phase zingg = ZinggWithSpark(args, options) zingg.initAndExecute() However, I am encountering the following error: ['--phase', 'findTrainingData'] arguments for client options are ['--phase', 'findTrainingData', '--license', 'zinggLic.txt', '--email', 'zingg@zingg.ai', '--conf', 'dummyConf.json'] AttributeError: 'list' object has no attribute 'getArgs' --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) File <command-5568433078295578>, line 4 1 options = ClientOptions([ClientOptions.PHASE,"findTrainingData"]) 3 #Zingg execution for the given phase ----> 4 zingg = ZinggWithSpark(args, options) 5 zingg.initAndExecute() File /local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.11/site-packages/zingg/client.py:370, in ZinggWithSpark.__init__(self, args, options) 369 def __init__(self, args, options): --> 370 self.client = getJVM().zingg.spark.client.SparkClient(args.getArgs(), options.getClientOptions(), getSparkSession()._jsparkSession) AttributeError: 'list' object has no attribute 'getArgs' Troubleshooting Steps Tried Verified that ClientOptions.PHASE and "findTrainingData" are correctly passed. Checked the Zingg documentation, but there are no clear examples regarding how ClientOptions should be used in this scenario. Tried explicitly defining args as a list with required options. Looked into the client.py implementation inside Zingg and found that the args parameter is expected to have a getArgs() method, but I am passing a simple list. Ensured that all required dependencies are installed and properly configured. Questions/Help Required How should args be passed correctly to ZinggWithSpark? Is ClientOptions being used correctly in this case, or is there an alternative approach?
