Closed erikerlandson closed 8 years ago
or is it more standard to use '.vectorid' ?
I have a more flexible and yet simpler way to do this:
case class DenseTaggedVector[+T](data: Array[Double], tag: T)
extends org.apache.spark.mllib.linalg.DenseVector(data)
case class SparseTaggedVector[+T](sz: Int, idx: Array[Int], data: Array[Double], tag: T)
extends org.apache.spark.mllib.linalg.SparseVector(sz, idx, data)
...an identifier payload