RANDOM CUT FOREST

I am trying to build a real time anomoly detection using kinesis data streams and reading data from streams in amazon managed flink and detecting anomoly using RCF using python wrapper

I am trying to run random cut forest using python in amazon managed flink

to load the jar files of RCF in python which i did locally

def start_jvm():
if not jpype.isJVMStarted():
lib_dir = Path(file).parent / "lib"
jar_path = str(lib_dir / "*")
# jar_path = "lib_2/randomcutforest-core-4.0.0-SNAPSHOT.jar:lib_2/randomcutforest-parkservices-4.0.0-SNAPSHOT.jar"
jpype.addClassPath(jar_path)
print(f"Starting JVM with classpath: {jar_path}")
jpype.startJVM(convertStrings=False)
print("JVM started successfully.")
else:
print("JVM already started.")

i need to stat jvm which is running file in my local .

but i am unable to run in amaon managed flink

how to achieve it .Thanks in advance

Start with the start_jvm() snippet and local lib JAR/classpath setup described in the issue. Compare that setup with the Amazon Managed Service for Apache Flink runtime and reproduce the JVM or classpath failure there. Done means documenting a working way to load the Random Cut Forest JARs and run the Python anomaly-detection flow.