Resets the internal variables.
Resets the internal variables. Called in the beginning of ´compute()´ of the RunRDD.
Updates hour and date and returns true if this sample belongs to a new bucket.
Updates hour and date and returns true if this sample belongs to a new bucket.
For the very first sample, returns false. Returns false in other cases.
a single sample of data type T.
For the very first sample, returns false. Returns true if this sample belongs to a new bucket. Returns false in other cases.
Simple class to be used with BucketRDD for detection of buckets from data such as http://www.cs.sandia.gov/~jrstear/logs/ by date and hour.
Usage:
import spark._ import spark.SparkContext._ import spark.timeseries._ val sc = TimeSeriesSpark.init("local[1]", "default") val file = sc.textFile("fileName") val mapped = file.map(_.split(" ", 8)) var det = new HourBucketDetector() val buckets = new BucketRDD(mapped, det) buckets.cache() val sizes = buckets.map(_.length).collect() for(k <- sizes) println(k)See spark.timeseries.BucketLogsByHour for a more real example.