NDArray

class NDArray(data, options)

Constructs or copies an NDArray instance.

Arguments:

• options (Object)

• options.shape (Array.<Number>)

• options.length (Number)

• options.strides (Array.<Number>)

• options.dtype (string)

Examples:

import { NDArray } from 'vectorious';

new NDArray() // => array([], dtype=float64)
new NDArray([]) // => array([], dtype=float64)
new NDArray([1, 2, 3]) // => array([1, 2, 3], dtype=float64)
new NDArray([[1, 2], [3, 4]]) // => array([ [ 1, 2 ], [ 3, 4 ] ], dtype=float64)
new NDArray(new Int32Array([1, 2, 3])) // => array([ 1, 2, 3 ], dtype=int32)
new NDArray([1, 2, 3, 4], {
  shape: [2, 2],
  dtype: 'uint32'
}) // => array([ [ 1, 2 ], [ 3, 4 ] ], dtype=uint32)

NDArray.T

type: NDArray

Short for this.copy().transpose()

NDArray.data

type: TypedArray

NDArray.dtype

type: String

NDArray.length

type: Number

NDArray.shape

type: Array.<Number>

NDArray.strides

type: Array.<Number>

NDArray.w

type: Number

Gets or sets the value at index 3

NDArray.x

type: Number

Gets or sets the value at index 0

NDArray.y

type: Number

Gets or sets the value at index 1

NDArray.z

type: Number

Gets or sets the value at index 2

NDArray.abs()

Returns the absolute value of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([-1, -2, -3]).abs() // <=> array([1, 2, 3])

NDArray.acos()

Returns the arccosine of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([-1, 0, 1]).acos(); // <=> array([3.141592653589793, 1.5707963267948966, 0])

NDArray.acosh(x)

Returns the hyperbolic arccosine of each element of current array.

Arguments:

• x (NDArray)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).acosh(); // <=> array([0, 1.316957950592041, 1.7627471685409546])

NDArray.add(x, 1)

Adds x multiplied by alpha to the current array. Accelerated with BLAS ?axpy.

Arguments:

• x (NDArray)

• 1 (Number) – alpha

Returns:

NDArray –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).add([4, 5, 6]); // <=> array([5, 7, 9])

NDArray.angle(x)

Determines the angle between the current vector and x.

Arguments:

• x (NDArray)

Returns:

number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).angle([4, 5, 6]); // <=> 0.22572622788897287

NDArray.asin()

Returns the arcsine of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([-1, 0, 1]).asin() // <=> array([-1.5707963705062866, 0, 1.5707963705062866])

NDArray.asinh()

Returns the hyperbolic arcsine of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([0, 1, 2]).asinh() // <=> array([0, 0.8813735842704773, 1.4436354637145996])

NDArray.atan()

Returns the arctangent of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).atan() // <=> array([0.7853981852531433, 1.1071487665176392, 1.249045729637146])

NDArray.atanh()

Returns the hyperbolic arctangent of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([0, -0.5]).atanh(); // <=> array([0, -0.5493061542510986])

NDArray.augment(x)

Augments x with current matrix.

Arguments:

• x (NDArray)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([[1, 2], [3, 4]]).augment(array([[1], [2]])); // <=> array([[1, 2, 1], [3, 4, 2]])

NDArray.binOp(x)

Perform binary operation f on x in the current array.

Arguments:

• x (NDArray)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).binOp([4, 5, 6], (a, b) => a + b); // => array([[5, 7, 9])

NDArray.cbrt()

Returns the cube root of each element of current array.

Returns:

this –

Examples:

import { cbrt } from 'vectorious';

cbrt([1, 8, 27]); // => array([1, 2, 3])

NDArray.ceil()

Returns smallest integer greater than or equal to of each element of current array.

Returns:

NDArray –

Examples:

import { array } from 'vectorious';

array([0.5, 1.5, 2.5]).ceil(); // <=> array([1, 2, 3])

NDArray.check()

Asserts if indices i, j, …, n are within the bounds of current array

Arguments:

• ...indices (Array.<Number>)

Throws:

Error – index out of bounds

Examples:

import { array } from 'vectorious';

array([0.5, 1.5, 2.5]).check(3); // Error: index out of bounds

NDArray.combine(x)

Combines the current vector with x

Arguments:

• x (NDArray)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).combine([4, 5, 6]); // => array([1, 2, 3, 4, 5, 6])

NDArray.copy()

Makes a copy of the class and underlying data

Returns:

NDArray –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).copy(); // => array([1, 2, 3])

NDArray.cos()

Returns the cosine of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([0, Math.PI / 2, Math.PI]).cos(); // => array([1, 0, -1])

NDArray.cosh()

Returns the hyperbolic cosine of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([0, 1, 2]).cosh(); // => array([1, 1.5430806875228882, 3.762195587158203])

NDArray.cross(x)

Computes the cross product of the current vector and the vector x This operation can only calculated for vectors with three components. Otherwise it throws an exception.

Arguments:

• x (NDArray)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).cross([4, 5, 6]); // <=> array([-3, 6, -3])

NDArray.det()

Gets the determinant of current matrix using LU factorization.

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([[0, 1], [2, 3]]).det(); // => -2

NDArray.diagonal()

Gets the diagonal of current matrix.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).diagonal(); // => array([1, 4])

NDArray.dot(x)

Performs dot multiplication with x and current array Accelerated with BLAS ?dot.

Arguments:

• x (NDArray)

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).dot([4, 5, 6]); // => 32

NDArray.eig()

Gets eigenvalues and eigenvectors of the current matrix using the Jacobi method. Accelerated with LAPACK ?geev.

Returns:

Array.<NDArray> –

Examples:

import { array } from 'vectorious';

array([[1, 0, 0], [0, 2, 0], [0, 0, 3]]).eig(); // => [array([1, 2, 3]), array([[1, 0, 0], [0, 1, 0], [0, 0, 1]])]

NDArray.equals(x, tolerance)

Checks if current array and x are equal.

Arguments:

• x (NDArray)

• tolerance (Number)

Returns:

Boolean –

Examples:

import { equals } from 'vectorious';

array([1, 2, 3]).equals([1, 2, 3]); // => true

NDArray.equidimensional(x)

Note

Deprecated.

Asserts if current array and x have the same shape

Arguments:

• x (NDArray)

Throws:

Error – shapes x and y do not match

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).equidimensional([1, 2]); // Error: shapes 3 and 2 do not match

NDArray.equilateral(x)

Note

Deprecated.

Asserts if current array and x have the same length

Arguments:

• x (NDArray)

Throws:

Error – lengths x and y do not match

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).equilateral([1, 2]); // Error: lengths 3 and 2 do not match

NDArray.exp()

Returns e^x of each element of current array, where x is the argument, and e is Euler’s constant (2.718…), the base of the natural logarithm.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).exp(); // <=> array([2.7182817459106445, 7.389056205749512, 20.08553695678711])

NDArray.expm1()

Returns subtracting 1 from exp(x) of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).expm1(); // <=> array([1.7182817459106445, 6.389056205749512, 19.08553695678711])

NDArray.fill(value)

Fills the current array with a scalar value

Arguments:

• value (Number)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).fill(0); // <=> array([0, 0, 0])

NDArray.floor()

Returns the largest integer less than or equal to a number of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1.5, 2.5, 3.5]).floor(); // <=> array([1, 2, 3])

NDArray.forEach(f)

Equivalent to TypedArray.prototype.forEach.

Arguments:

• f (function)

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).forEach(console.log);
// 1 0 [ 1, 2, 3 ]
// 2 1 [ 1, 2, 3 ]
// 3 2 [ 1, 2, 3 ]

NDArray.fround()

Returns the nearest single precision float representation of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([-5.05, 5.05]).fround(); // <=> array([-5.050000190734863, 5.050000190734863])

NDArray.gauss()

Gauss-Jordan elimination (i.e. returns the reduced row echelon form) of the current matrix.

Returns:

NDArray –

Examples:

import { array } from 'vectorious';

array([[1, 2, 3], [4, 5, 6]]).gauss(); // <=> array([[1, 0, -1], [-0, 1, 2]])

NDArray.get()

Gets the element at i, j, …, n from current vector.

Arguments:

• ...indices (Array.<Number>)

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).get(2); // 3

NDArray.inv()

Determines the inverse of current matrix using Gaussian elimination. Accelerated with LAPACK ?getri.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([[2, -1, 0], [-1, 2, -1], [0, -1, 2]]).inv(); // <=> array([[0.75, 0.5, 0.25], [0.5, 1, 0.5], [0.25, 0.5, 0.75]])

NDArray.log()

Returns the natural logarithm (log_e, also ln) of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).log(); // <=> array([0, 0.6931471824645996, 1.0986123085021973])

NDArray.log10()

Returns the base 10 logarithm of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([10, 100, 1000]).log10(); // <=> array([1, 2, 3])

NDArray.log1p()

Returns the natural logarithm (log_e, also ln) of 1 + x for each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]); // <=> array([0.6931471824645996, 1.0986123085021973, 1.3862943649291992])

NDArray.log2()

Returns the base 2 logarithm of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 4]).log2(); // => array([0, 1, 2])

NDArray.lu()

Performs full LU decomposition on current matrix. Accelerated with LAPACK ?getrf.

Returns:

Array.<(NDArray|Int32Array)> –

Examples:

import { array } from 'vectorious';

array([[1, 3, 5], [2, 4, 7], [1, 1, 0]]).lu(); // => [array([[1, 0, 0], [0.5, 1, 0], [0.5, -1, 1]]), array([[2, 4, 7], [0, 1, 1.5], [0, 0, -2]]), Int32Array([2, 2, 3])]

NDArray.lu_factor()

Performs LU factorization on current matrix. Accelerated with LAPACK ?getrf.

Returns:

Array.<(NDArray|Int32Array)> –

Examples:

import { array } from 'vectorious';

array([[1, 3, 5], [2, 4, 7], [1, 1, 0]]).lu_factor(); // <=> [array([[2, 4, 7], [0.5, 1, 1.5], [0.5, -1, -2]]), Int32Array([2, 2, 3])]

NDArray.map()

Equivalent to TypedArray.prototype.map.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).map(value => -value); // => array([-1, -2, -3])

NDArray.max()

Gets the maximum value (smallest) element of current array. Accelerated with BLAS i?amax.

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).max(); // => 3

NDArray.mean()

Gets the arithmetic mean of current array.

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).mean(); // => 2

NDArray.min()

Gets the minimum value (smallest) element of current array.

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).min(); // 1

NDArray.multiply(x)

Multiplies current matrix with x. Accelerated with BLAS ?gemm.

Arguments:

• x (NDArray)

Returns:

NDArray –

Examples:

import { array } from 'vectorious';

array([[1, 2]]).multiply([[1], [2]]); // <=> array([[5]])

NDArray.norm()

Calculates the norm of current array (also called L2 norm or Euclidean length). Accelerated with BLAS ?nrm2.

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).norm(); // => 3.7416574954986572

NDArray.normalize()

Normalizes current vector.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).normalize(); // => array([0.26726123690605164, 0.5345224738121033, 0.8017836809158325])

NDArray.pow(exponent)

Returns each element of current array to the exponent power, that is, element^exponent.

Arguments:

• exponent (Number)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).pow(2); // <=> array([1, 4, 9])

NDArray.prod()

Product of all elements of current array

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).prod(); // => 6

NDArray.product()

Hadamard product of current matrix and x

Returns:

NDArray –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).product([4, 5, 6]); // <=> array([4, 10, 18])

NDArray.project()

Projects the current vector onto x using the projection formula (y * (x * y / y * y)).

Returns:

NDArray –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).project([4, 5, 6]); // <=> array([1.6623376607894897, 2.0779221057891846, 2.49350643157959])

NDArray.push(value)

Pushes a new value into current vector.

Arguments:

• value (Number)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).push(4); // => array([1, 2, 3, 4])

NDArray.rank(tolerance)

Finds the rank of current matrix using gaussian elimination.

Arguments:

• tolerance (Number)

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([[1, 1, 1], [2, 2, 2], [3, 3, 3]]).rank(); // => 1

NDArray.reciprocal()

Gets the element-wise reciprocal of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]); // => array([1, 0.5, 0.3333333432674408])

NDArray.reduce(f, initialValue)

Equivalent to TypedArray.prototype.reduce.

Arguments:

• f (function)

• initialValue (Number)

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).reduce((a, b) => a + b, 0); // => 6

NDArray.reshape()

Reshapes current array

Arguments:

• ...shape (Array.<Number>)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3, 4]).reshape(2, 2); // <=> array([[1, 2], [3, 4]])

NDArray.round()

Returns the value of each element of current array rounded to the nearest integer.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1.2, 2.8, 3.5]).round(); // <=> array([1, 3, 4])

NDArray.row_add(dest, source, scalar)

Adds a multiple of one row multiplied by scalar to another inside current matrix.

Arguments:

• dest (Number)

• source (Number)

• scalar (Number)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([[1, 2], [3, 4]]).row_add(1, 0, 2); // <=> array([[1, 2], [5, 8]])

NDArray.scale(scalar)

Multiplies all elements of current array with a specified scalar. Accelerated with BLAS ?scal.

Arguments:

• scalar (Number)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).scale(2); // <=> array([2, 4, 6])

NDArray.set(value)

Sets the element at i, j, …, n to value.

Arguments:

• ...indices (Array.<Number>)

• value (Number)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).set(1, 0); // <=> array([1, 0, 3])

NDArray.sign()

Returns the sign of each element of current array, indicating whether it is positive, negative or zero.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).sign(); // <=> array([1, 1, 1])

NDArray.sin()

Returns the sine of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([0, Math.PI / 2, Math.PI]).sin(); // <=> array([0, 1, 0])

NDArray.sinh()

Returns the hyperbolic sine of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).sinh(); // <=> array([1.175201177597046, 3.6268603801727295, 10.017874717712402])

NDArray.slice(begin, end, step)

Slices the current array along the leading dimension

Arguments:

• begin (Number)

• end (Number)

• step (Number)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3, 4]).slice(0, 4, 2); // => array([1, 3])

NDArray.solve(x)

Solves the equation AX = B (where A is current matrix and B is x). Accelerated with LAPACK ?gesv.

Arguments:

• x (NDArray)

Returns:

NDArray –

Examples:

import { array } from 'vectorious';

array([[1, 3, 5], [2, 4, 7], [1, 1, 0]]).solve([[1], [3], [5]]); // => array([[3.25], [1.75], [-1.5]])

NDArray.sqrt()

Returns the positive square root of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 4, 9]); // <=> array([1, 2, 3])

NDArray.square()

Asserts if current matrix is square.

Throws:

Error – matrix is not square

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).square(); // Error: matrix is not square

NDArray.subtract()

Subtracts x from the current array. Accelerated with BLAS ?axpy.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).subtract([1, 1, 1]); // <=> array([0, 1, 2])

NDArray.sum()

Sum of array elements

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).sum(); // => 6

NDArray.swap(i, j)

Swaps two rows i and j in current matrix

Arguments:

• i (Number)

• j (Number)

Returns:

this –

Examples:

import { array } from 'vectorious';

array([[1, 2], [3, 4]]); // <=> array([[3, 4], [1, 2]])

NDArray.tan()

Returns the tangent of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).tan(); // <=> array([1.5574077367782593, -2.185039758682251, -0.14254654943943024])

NDArray.tanh()

Returns the hyperbolic tangent of each element of current array.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).tanh(); // <=> array([0.7615941762924194, 0.9640275835990906, 0.9950547814369202])

NDArray.toArray(index, dim)

Converts current vector into a JavaScript array.

Arguments:

• index (Number)

• dim (Number)

Returns:

Array –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).toArray(); // => [1, 2, 3]

NDArray.toString()

Converts current vector into a readable formatted string.

Returns:

String –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).toString(); // => '1,2,3'

NDArray.trace()

Gets the trace of the matrix (the sum of all diagonal elements).

Returns:

Number –

Examples:

import { array } from 'vectorious';

array([1, 2, 3]).trace(); // => 5

NDArray.transpose()

Transposes current matrix (mirror across the diagonal).

Returns:

this –

Examples:

import { array } from 'vectorious';

array([[1, 2, 3], [4, 5, 6], [7, 8, 9]]); // <=> array([[1, 4, 7], [2, 5, 8], [3, 6, 9]])

NDArray.trunc()

Returns the integer part of each element of current array, removing any fractional digits.

Returns:

this –

Examples:

import { array } from 'vectorious';

array([1.2, 2.8, 3.5]).trunc(); // => array([1, 2, 3])