Struct Statistics 

pub struct Statistics {
    pub samples: usize,
    pub min_us: f64,
    pub p50_us: f64,
    pub p90_us: f64,
    pub p95_us: f64,
    pub p99_us: f64,
    pub max_us: f64,
    pub mean_us: f64,
    pub stddev_us: f64,
}

The statistics of one set of samples, in microseconds.

Fields

samples: usize

How many samples.

min_us: f64

The smallest.

p50_us: f64

The median.

p90_us: f64

The 90th percentile.

p95_us: f64

The 95th percentile.

p99_us: f64

The 99th percentile.

max_us: f64

The largest.

mean_us: f64

The mean.

stddev_us: f64

The population standard deviation.

Implementations

impl Statistics

pub fn of(samples: &[Duration]) -> Self

Reduce samples. An empty set is all zeros with samples: 0.

use std::time::Duration;
use majordomus_cli::bench::Statistics;
let s = Statistics::of(&(1..=100).map(Duration::from_micros).collect::<Vec<_>>());
assert_eq!(s.samples, 100);
assert_eq!(s.min_us, 1.0);
assert_eq!(s.max_us, 100.0);
assert_eq!(s.p50_us, 50.0);
assert_eq!(s.p99_us, 99.0);
assert!((s.mean_us - 50.5).abs() < 1e-9);

pub fn samples_at_or_above(&self, metric: &str) -> Option<usize>

How many samples are certainly at or above a metric’s value: the samples a verdict on that metric rests on. A nearest-rank percentile of rank r has n - r + 1 samples at or above it; the maximum has one, the minimum all of them; the mean is an aggregate of every sample and one outlier can carry it, so it counts as one. None for a metric this does not know.

A metric over its threshold is a regression only when at least two samples agree: the median of one sample, or the p99 of twenty, is the single slowest sample, which one scheduler hiccup on a loaded runner decides.

use std::time::Duration;
use majordomus_cli::bench::Statistics;
let one = Statistics::of(&[Duration::from_millis(23)]);
assert_eq!(one.samples_at_or_above("p50"), Some(1));
let five = Statistics::of(&vec![Duration::from_millis(9); 5]);
assert_eq!(five.samples_at_or_above("p50"), Some(3));
assert_eq!(five.samples_at_or_above("p99"), Some(1));
assert_eq!(five.samples_at_or_above("min"), Some(5));
assert_eq!(five.samples_at_or_above("mean"), Some(1));
assert_eq!(five.samples_at_or_above("p42"), None);

pub fn metric(&self, name: &str) -> Option<f64>

One metric by name (p50, p95, p99, mean, max, min, p90).

Trait Implementations

impl Clone for Statistics

fn clone(&self) -> Statistics

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Statistics

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for Statistics

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl JsonSchema for Statistics

fn schema_name() -> Cow<'static, str>

The name of the generated JSON Schema.

fn schema_id() -> Cow<'static, str>

Returns a string that uniquely identifies the schema produced by this type.

fn json_schema(generator: &mut SchemaGenerator) -> Schema

Generates a JSON Schema for this type.

fn inline_schema() -> bool

Whether JSON Schemas generated for this type should be included directly in parent schemas, rather than being re-used where possible using the $ref keyword.

impl PartialEq for Statistics

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

Tests for self and other values to be equal, and is used by ==.

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

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for Statistics

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for Statistics