Enum Reducer 

pub enum Reducer {
    Count,
    Sum,
    Avg,
    Min,
    Max,
    First,
    Last,
    Median,
    Stdev,
    Percentile(f64),
    Difference,
    Keep,
    Unique,
    Top(usize),
    Samples,
}

A built-in reducer.

Variants

Count

Number of present (non-missing) values.

Sum

Sum of present values; 0 when there are none.

Avg

Mean of present values; missing when there are none. "mean" parses to this too.

Min

Smallest present value.

Max

Largest present value.

First

First present value.

Last

Last present value.

Median

50th percentile, interpolated.

Stdev

Population standard deviation (divides by n).

Percentile(f64)

Percentile q in [0, 100], linear interpolation between ranks ("p95").

Difference

Largest present value minus smallest (the spread); missing when there are none. Number and decimal columns only. In rolling it can be -0 (see crate::TimeSeries::rolling).

Keep

The value when every present value is equal, otherwise missing (also missing when there are none). Any column kind; the output has the source’s kind. Numbers compare as JavaScript’s === does, so 0 equals -0: aggregate and reduce give the first one, rolling gives 0. Array cells compare by their elements bit for bit (so [0] and [-0] differ), where pond-ts compares the array objects (so two equal arrays in different rows are kept here and give a gap in pond-ts).

Unique

The distinct present values as a list (an array column), sorted: numbers, then text, then true/false; text in JavaScript’s order (by UTF-16 code units). An array column’s lists are flattened one level first. 0 and -0 are one value, written 0. Nothing present gives an empty list, not a gap. Any column kind; decimals are read as their nearest f64.

Top(usize)

The N most frequent present values ("top3") as a list, most frequent first, ties in Reducer::Unique’s order; flattening, zeros and empty lists as there. N is at least 1.

Samples

Every present value as a list in arrival order, duplicates and -0 kept; flattening and empty lists as Reducer::Unique.

Implementations

impl Reducer

pub fn parse(name: &str) -> Result<Self, PondError>

Parse a pond-ts reducer name: count, sum, avg, mean, min, max, first, last, median, stdev, p<q>, difference, keep, unique, samples, top<N>.

Trait Implementations

impl Clone for Reducer

fn clone(&self) -> Self

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Copy for Reducer

impl Debug for Reducer

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl FromStr for Reducer

type Err = PondError

The associated error which can be returned from parsing.

fn from_str(name: &str) -> Result<Self, PondError>

Parses a string s to return a value of this type.

impl PartialEq for Reducer

fn eq(&self, other: &Self) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for Reducer