Trait rstats::Vecf64

pub trait Vecf64 {
    fn smult(self, s: f64) -> Vec<f64>;
    fn sadd(self, s: f64) -> Vec<f64>;
    fn dotp(self, v: &[f64]) -> f64;
    fn vinverse(self) -> Vec<f64>;
    fn cosine(self, _v: &[f64]) -> f64;
    fn vsub(self, v: &[f64]) -> Vec<f64>;
    fn negv(self) -> Vec<f64>;
    fn vadd(self, v: &[f64]) -> Vec<f64>;
    fn vmag(self) -> f64;
    fn vmagsq(self) -> f64;
    fn vdist(self, v: &[f64]) -> f64;
    fn vdistsq(self, v: &[f64]) -> f64;
    fn vunit(self) -> Vec<f64>;
    fn varea(self, v: &[f64]) -> f64;
    fn varc(self, v: &[f64]) -> f64;
    fn vsim(self, v: &[f64]) -> f64;
    fn vdisim(self, v: &[f64]) -> f64;
    fn correlation(self, _v: &[f64]) -> f64;
    fn kendalcorr(self, _v: &[f64]) -> f64;
    fn spearmancorr(self, _v: &[f64]) -> f64;
    fn kazutsugi(self) -> f64;
    fn autocorr(self) -> f64;
    fn minmax(self) -> (f64, usize, f64, usize);
    fn lintrans(self) -> Vec<f64>;
    fn binsearch(self, v: f64) -> usize;
    fn merge(self, v: &[f64]) -> Vec<f64>;
    fn merge_immutable(
        self, 
        idx1: &[usize], 
        v2: &[f64], 
        idx2: &[usize]
    ) -> (Vec<f64>, Vec<usize>);
    fn merge_indices(self, idx1: &[usize], idx2: &[usize]) -> Vec<usize>;
    fn sortf(self) -> Vec<f64>;
    fn sortm(self, ascending: bool) -> Vec<f64>;
    fn mergerank(self) -> Vec<usize>;
    fn mergesort(self, i: usize, n: usize) -> Vec<usize>;
}

Vector algebra on one or two vectors.

Required methods

fn smult(self, s: f64) -> Vec<f64>

Scalar multiplication of a vector

fn sadd(self, s: f64) -> Vec<f64>

Scalar addition to vector

fn dotp(self, v: &[f64]) -> f64

Scalar product of two vectors

fn vinverse(self) -> Vec<f64>

Inverse vecor of magnitude 1/|v|

fn cosine(self, _v: &[f64]) -> f64

Cosine = a.dotp(b)/(a.vmag*b.vmag)

fn vsub(self, v: &[f64]) -> Vec<f64>

Vector subtraction

fn negv(self) -> Vec<f64>

Vector negtion

fn vadd(self, v: &[f64]) -> Vec<f64>

Vector addition

fn vmag(self) -> f64

Vector magnitude

fn vmagsq(self) -> f64

Vector magnitude squared

fn vdist(self, v: &[f64]) -> f64

Euclidian distance between two points

fn vdistsq(self, v: &[f64]) -> f64

vdist between two points squared

fn vunit(self) -> Vec<f64>

Unit vector

fn varea(self, v: &[f64]) -> f64

Area of parallelogram between two vectors (magnitude of cross product)

fn varc(self, v: &[f64]) -> f64

Area proportional to the swept arc

fn vsim(self, v: &[f64]) -> f64

Vector similarity in the interval [0,1]: (1+cos(theta))/2

fn vdisim(self, v: &[f64]) -> f64

Vector dissimilarity in the interval [0,1]: (1-cos(theta))/2

fn correlation(self, _v: &[f64]) -> f64

Correlation

fn kendalcorr(self, _v: &[f64]) -> f64

Kendall’s tau-b (rank order) correlation

fn spearmancorr(self, _v: &[f64]) -> f64

Spearman’s rho (rank differences) correlation

fn kazutsugi(self) -> f64

Kazutsugi Spearman’s corelation against just five distances (to outcomes classes)

fn autocorr(self) -> f64

Autocorrelation

fn minmax(self) -> (f64, usize, f64, usize)

Minimum, minimum’s index, maximum, maximum’s index.

fn lintrans(self) -> Vec<f64>

Linear transformation to [0,1]

fn binsearch(self, v: f64) -> usize

Binary search for insert index I in sorted vector

fn merge(self, v: &[f64]) -> Vec<f64>

Merges two ascending sorted vectors

fn merge_immutable(
    self,
    idx1: &[usize],
    v2: &[f64],
    idx2: &[usize]
) -> (Vec<f64>, Vec<usize>)

Merges two sort indices, returns simply concatenated Vec and new sort index into it

fn merge_indices(self, idx1: &[usize], idx2: &[usize]) -> Vec<usize>

merge indices of two already concatenated sorted vectors

fn sortf(self) -> Vec<f64>

Sort vector in a standard way

fn sortm(self, ascending: bool) -> Vec<f64>

Sorted vector, is wrapper for mergesort below

fn mergerank(self) -> Vec<usize>

Ranking with only n*log(n) complexity, using ‘mergesort’

fn mergesort(self, i: usize, n: usize) -> Vec<usize>

Immutable merge sort, makes a sort index

Implementations on Foreign Types

impl Vecf64 for &[f64]

fn smult(self, s: f64) -> Vec<f64>

Scalar multiplication of a vector, creates new vec

fn sadd(self, s: f64) -> Vec<f64>

Scalar addition to a vector, creates new vec

fn dotp(self, v: &[f64]) -> f64

Scalar product of two f64 slices.
Must be of the same length - no error checking (for speed)

fn cosine(self, v: &[f64]) -> f64

Cosine of an angle between two vectors.

Example

use rstats::Vecf64;
let v1 = vec![1_f64,2.,3.,4.,5.,6.,7.,8.,9.,10.,11.,12.,13.,14.];
let v2 = vec![14_f64,1.,13.,2.,12.,3.,11.,4.,10.,5.,9.,6.,8.,7.];
assert_eq!(v1.cosine(&v2),0.7517241379310344);

fn vsub(self, v: &[f64]) -> Vec<f64>

Vector subtraction, creates a new Vec result

fn vadd(self, v: &[f64]) -> Vec<f64>

Vector addition, creates a new Vec result

fn vdist(self, v: &[f64]) -> f64

Euclidian distance between two n dimensional points (vectors).
Slightly faster than vsub followed by vmag, as both are done in one loop

fn vdistsq(self, v: &[f64]) -> f64

Euclidian distance squared between two n dimensional points (vectors).
Slightly faster than vsub followed by vmasq, as both are done in one loop Same as vdist without taking the square root

fn vmag(self) -> f64

Vector magnitude

fn vmagsq(self) -> f64

Vector magnitude squared

fn vunit(self) -> Vec<f64>

Unit vector - creates a new one

fn varea(self, v: &[f64]) -> f64

Area of a parallelogram between two vectors. Same as the magnitude of their cross product |a ^ b| = |a||b|sin(theta). Attains maximum |a|.|b| when the vectors are othogonal.

fn varc(self, v: &[f64]) -> f64

Area proportional to the swept arc up to angle theta. Attains maximum of 2|a||b| when the vectors have opposite orientations. This is really |a||b|(1-cos(theta)) = 2|a||b|D

fn vsim(self, v: &[f64]) -> f64

We define vector similarity S in the interval [0,1] as S = (1+cos(theta))/2

fn vdisim(self, v: &[f64]) -> f64

We define vector dissimilarity D in the interval [0,1] as D = 1-S = (1-cos(theta))/2

fn correlation(self, v: &[f64]) -> f64

Pearson’s correlation coefficient of a sample of two f64 variables.

Example

use rstats::Vecf64;
let v1 = vec![1_f64,2.,3.,4.,5.,6.,7.,8.,9.,10.,11.,12.,13.,14.];
let v2 = vec![14_f64,1.,13.,2.,12.,3.,11.,4.,10.,5.,9.,6.,8.,7.];
assert_eq!(v1.correlation(&v2),-0.1076923076923077);

fn kendalcorr(self, v: &[f64]) -> f64

Kendall Tau-B correlation coefficient of a sample of two f64 variables. Defined by: tau = (conc - disc) / sqrt((conc + disc + tiesx) * (conc + disc + tiesy)) This is the simplest implementation with no sorting.

Example

use rstats::Vecf64;
let v1 = vec![1_f64,2.,3.,4.,5.,6.,7.,8.,9.,10.,11.,12.,13.,14.];
let v2 = vec![14_f64,1.,13.,2.,12.,3.,11.,4.,10.,5.,9.,6.,8.,7.];
assert_eq!(v1.kendalcorr(&v2),-0.07692307692307693);

fn spearmancorr(self, v: &[f64]) -> f64

Spearman rho correlation coefficient of two f64 variables. This is the simplest implementation with no sorting.

Example

use rstats::Vecf64;
let v1 = vec![1_f64,2.,3.,4.,5.,6.,7.,8.,9.,10.,11.,12.,13.,14.];
let v2 = vec![14_f64,1.,13.,2.,12.,3.,11.,4.,10.,5.,9.,6.,8.,7.];
assert_eq!(v1.spearmancorr(&v2),-0.1076923076923077);

fn kazutsugi(self) -> f64

Spearman correlation of five distances against Kazutsugi discrete outcomes [0.00,0.25,0.50,0.75,1.00], ranked as [4,3,2,1,0] (the order is swapped to penalise distances). The result is in the range [-1,1].

Example

use rstats::Vecf64;
let v1:Vec<f64> = vec![4.,1.,2.,0.,3.];
assert_eq!(v1.kazutsugi(),0.3);

fn autocorr(self) -> f64

(Auto)correlation coefficient of pairs of successive values of (time series) f64 variable.

Example

use rstats::Vecf64;
let v1 = vec![1_f64,2.,3.,4.,5.,6.,7.,8.,9.,10.,11.,12.,13.,14.];
assert_eq!(v1.autocorr(),0.9984603532054123_f64);

fn minmax(self) -> (f64, usize, f64, usize)

Finds minimum, minimum’s index, maximum, maximum’s index of &f64 Here self is usually some data, rather than a vector

fn lintrans(self) -> Vec<f64>

Linear transform to interval [0,1]

fn binsearch(self, v: f64) -> usize

Returns index to the first item that is strictly greater than v, using binary search of an ascending sorted list. When none are greater, returns self.len(). User must check for this index overflow: if the returned index == 0, then v was below the list, else use index-1 as a valid index to the last item that is less than or equal to v. This then is the right index to use for looking up cummulative probability density functions.

fn merge(self, v: &[f64]) -> Vec<f64>

Merges two ascending sorted vectors &f64

fn merge_immutable(
    self,
    idx1: &[usize],
    v2: &[f64],
    idx2: &[usize]
) -> (Vec<f64>, Vec<usize>)

Merges two ascending sorted vectors’ indices, returns concatenated Vec and new index into it. Mostly just a wrapper for merge_indices()

fn merge_indices(self, idx1: &[usize], idx2: &[usize]) -> Vec<usize>

Merges indices of two already concatenated sorted vectors: self is untouched, only sort indices are merged. Used by mergesort and merge_immutable.

fn sortm(self, ascending: bool) -> Vec<f64>

Immutable sort. Returns new sorted vector, just like ‘sortf’ above but using our indexing ‘mergesort’ below. Simply passes the boolean flag ‘ascending’ onto ‘unindex’.

fn mergerank(self) -> Vec<usize>

Ranking of self by inverting the (merge) sort index.
Sort index is in sorted order, giving indices to the original data positions. Ranking is in original data order, giving their positions in the sort index. Thus they are in an inverse relationship, easily converted by .revindex() Very fast ranking of many f64 items, ranking self with only n*(log(n)+1) complexity.

fn mergesort(self, i: usize, n: usize) -> Vec<usize>

Doubly recursive non-destructive merge sort. The data is read-only, it is not moved or mutated. Returns vector of indices to self from i to i+n, such that the indexed values are in sort order.
Thus we are moving only the index (key) values instead of the actual values.

fn sortf(self) -> Vec<f64>

New sorted vector. Immutable sort. Copies self and then sorts it in place, leaving self unchanged. Calls mutsortf and that calls the standard self.sort_unstable_by. Consider using our sortm instead.

fn vinverse(self) -> Vec<f64>

fn negv(self) -> Vec<f64>

Implementors