Trying to Test

[darkfrog26]

I've created a fork of this project to upgrade to the latest version of Spark and make a few other changes. However, I'm having problems understanding the simple use-case of working with this in order to create unit tests:

https://github.com/darkfrog26/pu4spark/blob/master/src/test/scala/specs/SimpleUsageSpec.scala#L24

Any feedback on proper operation of would be appreciated. I'm still a bit new to Spark ML, but your README gives very little clarity on proper operation.

[astrakhantsev]

val weightedDF = puLearner.weight(training, "label", "features")
// TODO: what's next?

Next you can use 'outputLabel' column of weightedDF dataframe - it would contain or each instance the number from 0 to 1 reflecting classifier's confidence for that instance, i.e. how likely this instance is positive or negative.