Github user rxin commented on a diff in the pull request:
https://github.com/apache/spark/pull/2455#discussion_r18122187
--- Diff:
core/src/main/scala/org/apache/spark/util/random/RandomSampler.scala ---
@@ -53,56 +81,237 @@ trait RandomSampler[T, U] extends Pseudorandom with
Cloneable with Serializable
* @tparam T item type
*/
@DeveloperApi
-class BernoulliSampler[T](lb: Double, ub: Double, complement: Boolean =
false)
+class BernoulliPartitionSampler[T](lb: Double, ub: Double, complement:
Boolean = false)
extends RandomSampler[T, T] {
- private[random] var rng: Random = new XORShiftRandom
+ // epsilon slop to avoid failure from floating point jitter
+ @transient val eps: Double = RandomSampler.epsArgs
+ require(lb <= (ub + eps), "Lower bound (lb) must be <= upper bound (ub)")
+ require(lb >= (0d - eps), "Lower bound (lb) must be >= 0.0")
+ require(ub <= (1d + eps), "Upper bound (ub) must be <= 1.0")
- def this(ratio: Double) = this(0.0d, ratio)
+ private val rng: Random = new XORShiftRandom
override def setSeed(seed: Long) = rng.setSeed(seed)
override def sample(items: Iterator[T]): Iterator[T] = {
- items.filter { item =>
- val x = rng.nextDouble()
- (x >= lb && x < ub) ^ complement
+ ub-lb match {
+ case f if (f <= 0d) => if (complement) items else Iterator.empty
+ case _ => {
+ if (complement) {
+ items.filter(item => {
+ val x = rng.nextDouble()
+ (x < lb) || (x >= ub)
+ })
+ } else {
+ items.filter(item => {
+ val x = rng.nextDouble()
+ (x >= lb) && (x < ub)
+ })
+ }
+ }
}
}
/**
* Return a sampler that is the complement of the range specified of
the current sampler.
*/
- def cloneComplement(): BernoulliSampler[T] = new BernoulliSampler[T](lb,
ub, !complement)
+ def cloneComplement(): BernoulliPartitionSampler[T] =
+ new BernoulliPartitionSampler[T](lb, ub, !complement)
- override def clone = new BernoulliSampler[T](lb, ub, complement)
+ override def clone = new BernoulliPartitionSampler[T](lb, ub, complement)
}
+
/**
* :: DeveloperApi ::
- * A sampler based on values drawn from Poisson distribution.
+ * A sampler based on Bernoulli trials.
*
- * @param mean Poisson mean
+ * @param fraction the sampling fraction, aka Bernoulli sampling
probability
* @tparam T item type
*/
@DeveloperApi
-class PoissonSampler[T](mean: Double) extends RandomSampler[T, T] {
+class BernoulliSampler[T: ClassTag](fraction: Double) extends
RandomSampler[T, T] {
+
+ // epsilon slop to avoid failure from floating point jitter
+ @transient val eps: Double = RandomSampler.epsArgs
+ require(fraction >= (0.0 - eps) && fraction <= (1.0 + eps),
+ "Sampling fraction must be on interval [0, 1]")
+
+ private val rng: Random = RandomSampler.rngDefault
- private[random] var rng = new Poisson(mean, new DRand)
+ override def setSeed(seed: Long) = rng.setSeed(seed)
+
+ override def sample(items: Iterator[T]): Iterator[T] = {
+ fraction match {
--- End diff --
again directly doing if else if probably more clear than pattern matching
... especially when pattern matching doesn't buy you anything syntactically here
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
