How can I serialize the object RandomCutForest to array bytes?

[zuoxiang95]

hey guys, did there have any solution to serialize RandomCutForest? Here is my serialize function, but it doesn't work:

  def serialize[T](o: T): Array[Byte] = {
    val bos = new ByteArrayOutputStream() 
    val oos = new ObjectOutputStream(bos)
    oos.writeObject(o)
    oos.close()
    bos.toByteArray
  }

[sudiptoguha]

First, take a look at https://github.com/aws/random-cut-forest-by-aws/blob/main/Java/examples/src/main/java/com/amazon/randomcutforest/examples/serialization/ObjectStreamExample.java

You can choose to save tree state/etc. There are options for Jackson. However bytes[] seem to be slow and perhaps you'd be (comparability) better served with ProtoStuff

https://github.com/aws/random-cut-forest-by-aws/blob/main/Java/examples/src/main/java/com/amazon/randomcutforest/examples/serialization/ProtostuffExampleWithShingles.java

However FYI as we push to the next release 4.0 we will likely be changing the state classes. In the next few PRs (in the next 3.X) we will update the layout of the information without breaking the state classes. As a philosophy, the library will focus on the functional aspects of decision forests and would not be able to add to new specific serialization formats.

But there is a sub-package https://github.com/aws/random-cut-forest-by-aws/tree/main/Java/serialization/src/main/java/com/amazon/randomcutforest/serialize and contributions/discussions are welcome. So far that package has focused on conversion for the major upgrades.

[zuoxiang95]

@sudiptoguha Thanks for your kindly reply. Your solution perfectly solved my problem!!!