stats

Type Members

1. case class Runs[K, V](elems: Seq[(K, V)])(implicit evidence$1: Integral[V]) extends Product with Serializable

   

   Convenience class wrapping a sequence of key-number pairs, used in run-length-encoding in Stats.

2. case class Samples[K, V](n: V, first: Runs[K, V], numFirst: V, last: Runs[K, V], numLast: V)(implicit evidence$1: Integral[V]) extends Product with Serializable

   

   Used by Stats to wrap some Runs of elements from the start and end of a dataset.

   Used by Stats to wrap some Runs of elements from the start and end of a dataset.

   K

   arbitrary element type

   V

   Integral type, e.g. Int or Long.

   n

   total number of elements in the dataset.

   first

   Runs of elements from the start of the dataset.

   numFirst

   the number of elements represented by the Runs in first, i.e. the sum of the their values.

   last

   Runs of elements from the end of the dataset.

   numLast

   the number of elements represented by the Runs in last, i.e. the sum of the their values.

3. case class Stats[K, V](n: V, mean: Double, stddev: Double, mad: Double, samplesOpt: Option[Samples[K, V]], sortedSamplesOpt: Option[Samples[K, V]], percentiles: Seq[(Double, Double)])(implicit evidence$1: Numeric[K], evidence$2: Integral[V]) extends Product with Serializable

   

   Wrapper for some computed statistics about a dataset of Numeric elements.

   Wrapper for some computed statistics about a dataset of Numeric elements.

   K

   Numeric element type. TODO(ryan): allow this to be non-Numeric.

   V

   Integral value type.

   n

   number of elements in the dataset.

   mean

   mean.

   stddev

   stddev.

   mad

   median absolute deviation (from the median).

   samplesOpt

   "sample" elements; the start and end of the data.

   sortedSamplesOpt

   "sample" elements; the least and greatest elements. If the dataset is already sorted, meaning this would be equivalent to samplesOpt, it is omitted.

   percentiles

   selected percentiles of the dataset.