NDArray Methods

The NDArray class (from the default numpy-ts entry point) provides chainable instance methods that delegate to the corresponding standalone functions. Each method below calls its standalone equivalent under the hood.

import * as np from 'numpy-ts';

// Method chaining
const result = np.arange(12)
  .reshape(3, 4)
  .multiply(2)
  .sum(0);

// Equivalent standalone calls
const a = np.arange(12);
const b = np.reshape(a, [3, 4]);
const c = np.multiply(b, 2);
const d = np.sum(c, 0);

These methods are only available on NDArray (the full entry point). NDArrayCore from numpy-ts/core does not include them. See the NDArray Class page for core methods available on both classes.

Arithmetic methods

add

add(other: NDArray | number): NDArray

Element-wise addition. See add.

subtract

subtract(other: NDArray | number): NDArray

Element-wise subtraction. See subtract.

multiply

multiply(other: NDArray | number): NDArray

Element-wise multiplication. See multiply.

divide

divide(other: NDArray | number): NDArray

Element-wise division. See divide.

floor_divide

floor_divide(other: NDArray | number): NDArray

Element-wise floor division. See floor_divide.

mod

mod(other: NDArray | number): NDArray

Element-wise modulo. See mod.

remainder

remainder(divisor: NDArray | number): NDArray

Element-wise remainder (same as mod). See remainder.

fmod

fmod(divisor: NDArray | number): NDArray

Element-wise C-style remainder. See fmod.

power

power(exponent: NDArray | number): NDArray

Raise elements to the given power. See power.

negative

negative(): NDArray

Element-wise numerical negation (-x). See negative.

positive

positive(): NDArray

Element-wise numerical positive (+x). Returns a copy. See positive.

absolute

absolute(): NDArray

Element-wise absolute value. See absolute.

reciprocal

reciprocal(): NDArray

Element-wise reciprocal (1/x). See reciprocal.

square

square(): NDArray

Element-wise square (x**2). See square.

sqrt

sqrt(): NDArray

Element-wise square root. See sqrt.

cbrt

cbrt(): NDArray

Element-wise cube root. See cbrt.

fabs

fabs(): NDArray

Element-wise absolute value (always returns float). See fabs.

sign

sign(): NDArray

Element-wise sign indication (-1, 0, or 1). See sign.

heaviside

heaviside(x2: NDArray | number): NDArray

Heaviside step function. See heaviside.

divmod

divmod(divisor: NDArray | number): [NDArray, NDArray]

Returns both floor quotient and remainder. See divmod.

Exponential and logarithmic methods

exp

exp(): NDArray

Natural exponential (e^x). See exp.

exp2

exp2(): NDArray

Base-2 exponential (2^x). See exp2.

expm1

expm1(): NDArray

e^x - 1, accurate for small x. See expm1.

log

log(): NDArray

Natural logarithm. See log.

log2

log2(): NDArray

Base-2 logarithm. See log2.

log10

log10(): NDArray

Base-10 logarithm. See log10.

log1p

log1p(): NDArray

ln(1 + x), accurate for small x. See log1p.

logaddexp

logaddexp(x2: NDArray | number): NDArray

log(exp(x1) + exp(x2)), numerically stable. See logaddexp.

logaddexp2

logaddexp2(x2: NDArray | number): NDArray

log2(2^x1 + 2^x2), numerically stable. See logaddexp2.

Trigonometric methods

sin

sin(): NDArray

Sine (radians). See sin.

cos

cos(): NDArray

Cosine (radians). See cos.

tan

tan(): NDArray

Tangent (radians). See tan.

arcsin

arcsin(): NDArray

Inverse sine. See arcsin.

arccos

arccos(): NDArray

Inverse cosine. See arccos.

arctan

arctan(): NDArray

Inverse tangent. See arctan.

arctan2

arctan2(other: NDArray | number): NDArray

Element-wise arc tangent of this/other, choosing the correct quadrant. See arctan2.

hypot

hypot(other: NDArray | number): NDArray

Hypotenuse given two right-triangle legs. See hypot.

degrees

degrees(): NDArray

Convert radians to degrees. See degrees.

radians

radians(): NDArray

Convert degrees to radians. See radians.

Hyperbolic methods

sinh

sinh(): NDArray

Hyperbolic sine. See sinh.

cosh

cosh(): NDArray

Hyperbolic cosine. See cosh.

tanh

tanh(): NDArray

Hyperbolic tangent. See tanh.

arcsinh

arcsinh(): NDArray

Inverse hyperbolic sine. See arcsinh.

arccosh

arccosh(): NDArray

Inverse hyperbolic cosine. See arccosh.

arctanh

arctanh(): NDArray

Inverse hyperbolic tangent. See arctanh.

Rounding methods

ceil

ceil(): NDArray

Ceiling of each element. See ceil.

floor

floor(): NDArray

Floor of each element. See floor.

fix

fix(): NDArray

Round towards zero. See fix.

rint

rint(): NDArray

Round to nearest integer. See rint.

trunc

trunc(): NDArray

Truncate to integer part. See trunc.

around

around(decimals?: number): NDArray

Round to the given number of decimals (default: 0). See around.

round

round(decimals?: number): NDArray

Alias for around. See around.

clip

clip(a_min: number | NDArray | null, a_max: number | NDArray | null): NDArray

Clip values to [a_min, a_max]. Pass null for no bound on either side. See clip.

Comparison methods

equal

equal(other: NDArray | number): NDArray

Element-wise equality. Returns boolean array. See equal.

not_equal

not_equal(other: NDArray | number): NDArray

Element-wise inequality. Returns boolean array. See not_equal.

greater

greater(other: NDArray | number): NDArray

Element-wise greater than. Returns boolean array. See greater.

greater_equal

greater_equal(other: NDArray | number): NDArray

Element-wise greater than or equal. Returns boolean array. See greater_equal.

less

less(other: NDArray | number): NDArray

Element-wise less than. Returns boolean array. See less.

less_equal

less_equal(other: NDArray | number): NDArray

Element-wise less than or equal. Returns boolean array. See less_equal.

allclose

allclose(other: NDArray | number, rtol?: number, atol?: number): boolean

Returns true if all elements are equal within tolerance. Default rtol=1e-5, atol=1e-8. See allclose.

isclose

isclose(other: NDArray | number, rtol?: number, atol?: number): NDArray

Element-wise comparison within tolerance. Returns boolean array. See isclose.

isnan

isnan(): NDArray

Test element-wise for NaN. Returns boolean array. See isnan.

isinf

isinf(): NDArray

Test element-wise for infinity. Returns boolean array. See isinf.

isfinite

isfinite(): NDArray

Test element-wise for finiteness. Returns boolean array. See isfinite.

isnat

isnat(): NDArray

Test element-wise for NaT (Not a Time). Always returns false (no datetime support). See isnat.

Logic methods

logical_and

logical_and(other: NDArray | number): NDArray

Element-wise logical AND. Returns boolean array. See logical_and.

logical_or

logical_or(other: NDArray | number): NDArray

Element-wise logical OR. Returns boolean array. See logical_or.

logical_not

logical_not(): NDArray

Element-wise logical NOT. Returns boolean array. See logical_not.

logical_xor

logical_xor(other: NDArray | number): NDArray

Element-wise logical XOR. Returns boolean array. See logical_xor.

invert

invert(): NDArray

Bitwise NOT (alias for bitwise_not). See invert.

bitwise_and

bitwise_and(other: NDArray | number): NDArray

Element-wise bitwise AND. Requires integer dtype. See bitwise_and.

bitwise_or

bitwise_or(other: NDArray | number): NDArray

Element-wise bitwise OR. Requires integer dtype. See bitwise_or.

bitwise_xor

bitwise_xor(other: NDArray | number): NDArray

Element-wise bitwise XOR. Requires integer dtype. See bitwise_xor.

bitwise_not

bitwise_not(): NDArray

Element-wise bitwise NOT. Requires integer dtype. See bitwise_not.

left_shift

left_shift(shift: NDArray | number): NDArray

Element-wise left bit shift. See left_shift.

right_shift

right_shift(shift: NDArray | number): NDArray

Element-wise right bit shift. See right_shift.

copysign

copysign(x2: NDArray | number): NDArray

Change the sign of this to that of x2, element-wise. See copysign.

signbit

signbit(): NDArray

Returns true where the sign bit is set (negative). See signbit.

nextafter

nextafter(x2: NDArray | number): NDArray

Return the next floating-point value after this towards x2. See nextafter.

spacing

spacing(): NDArray

Return the distance between each element and the nearest adjacent number. See spacing.

Reduction methods

sum

sum(axis?: number, keepdims?: boolean): NDArray | number | bigint | Complex

Sum of elements along an axis. Returns a scalar when axis is undefined. See sum.

prod

prod(axis?: number, keepdims?: boolean): NDArray | number | bigint | Complex

Product of elements along an axis. See prod.

mean

mean(axis?: number, keepdims?: boolean): NDArray | number | Complex

Arithmetic mean along an axis. See mean.

std

std(axis?: number, ddof?: number, keepdims?: boolean): NDArray | number

Standard deviation along an axis. ddof defaults to 0. See std.

var

var(axis?: number, ddof?: number, keepdims?: boolean): NDArray | number

Variance along an axis. ddof defaults to 0. See var.

min

min(axis?: number, keepdims?: boolean): NDArray | number | Complex

Minimum value along an axis. See min.

max

max(axis?: number, keepdims?: boolean): NDArray | number | Complex

Maximum value along an axis. See max.

argmin

argmin(axis?: number): NDArray | number

Index of the minimum value along an axis. See argmin.

argmax

argmax(axis?: number): NDArray | number

Index of the maximum value along an axis. See argmax.

all

all(axis?: number, keepdims?: boolean): NDArray | boolean

Test whether all elements evaluate to true. See all.

any

any(axis?: number, keepdims?: boolean): NDArray | boolean

Test whether any element evaluates to true. See any.

ptp

ptp(axis?: number, keepdims?: boolean): NDArray | number | Complex

Peak-to-peak (maximum minus minimum) along an axis. See ptp.

median

median(axis?: number, keepdims?: boolean): NDArray | number

Median along an axis. See median.

percentile

percentile(q: number, axis?: number, keepdims?: boolean): NDArray | number

Compute the q-th percentile (0—100). See percentile.

quantile

quantile(q: number, axis?: number, keepdims?: boolean): NDArray | number

Compute the q-th quantile (0—1). See quantile.

average

average(weights?: NDArray, axis?: number): NDArray | number | Complex

Weighted average along an axis. See average.

cumsum

cumsum(axis?: number): NDArray

Cumulative sum along an axis. See cumsum.

cumprod

cumprod(axis?: number): NDArray

Cumulative product along an axis. See cumprod.

diff

diff(n?: number, axis?: number): NDArray

Calculate the n-th discrete difference along an axis. See diff.

nansum

nansum(axis?: number, keepdims?: boolean): NDArray | number | Complex

Sum treating NaNs as zero. See nansum.

nanprod

nanprod(axis?: number, keepdims?: boolean): NDArray | number | Complex

Product treating NaNs as one. See nanprod.

nanmean

nanmean(axis?: number, keepdims?: boolean): NDArray | number | Complex

Mean ignoring NaNs. See nanmean.

nanstd

nanstd(axis?: number, ddof?: number, keepdims?: boolean): NDArray | number

Standard deviation ignoring NaNs. See nanstd.

nanvar

nanvar(axis?: number, ddof?: number, keepdims?: boolean): NDArray | number

Variance ignoring NaNs. See nanvar.

nanmin

nanmin(axis?: number, keepdims?: boolean): NDArray | number | Complex

Minimum ignoring NaNs. See nanmin.

nanmax

nanmax(axis?: number, keepdims?: boolean): NDArray | number | Complex

Maximum ignoring NaNs. See nanmax.

nanargmin

nanargmin(axis?: number): NDArray | number

Index of minimum ignoring NaNs. See nanargmin.

nanargmax

nanargmax(axis?: number): NDArray | number

Index of maximum ignoring NaNs. See nanargmax.

nanmedian

nanmedian(axis?: number, keepdims?: boolean): NDArray | number

Median ignoring NaNs. See nanmedian.

nancumsum

nancumsum(axis?: number): NDArray

Cumulative sum treating NaNs as zero. See nancumsum.

nancumprod

nancumprod(axis?: number): NDArray

Cumulative product treating NaNs as one. See nancumprod.

nanpercentile

nanpercentile(q: number, axis?: number, keepdims?: boolean): NDArray | number

Percentile ignoring NaNs. See nanpercentile.

nanquantile

nanquantile(q: number, axis?: number, keepdims?: boolean): NDArray | number

Quantile ignoring NaNs. See nanquantile.

Shape methods

reshape

reshape(...shape: number[]): NDArray

Return a new array with the given shape. Use -1 for one inferred dimension. See reshape.

flatten

flatten(): NDArray

Return a flattened copy (always 1D). See flatten.

ravel

ravel(): NDArray

Return a flattened array. Returns a view when possible, otherwise a copy. See ravel.

squeeze

squeeze(axis?: number): NDArray

Remove axes of length one. See squeeze.

expand_dims

expand_dims(axis: number): NDArray

Insert a new axis of length one at the given position. See expand_dims.

transpose

transpose(axes?: number[]): NDArray

Permute the dimensions. With no arguments, reverses all axes. See transpose.

swapaxes

swapaxes(axis1: number, axis2: number): NDArray

Interchange two axes. See swapaxes.

moveaxis

moveaxis(source: number | number[], destination: number | number[]): NDArray

Move axes to new positions. See moveaxis.

T

get T(): NDArray

Transpose property (no parentheses). Returns a view with reversed axes. See NDArray Class.

Linear algebra methods

dot

dot(other: NDArray): NDArray | number | bigint | Complex

Dot product following NumPy semantics (inner product for 1D, matrix multiply for 2D). See dot.

matmul

matmul(other: NDArray): NDArray

Matrix multiplication (@ operator equivalent). See matmul.

inner

inner(other: NDArray): NDArray | number | bigint | Complex

Inner product (contracts over last axes of both arrays). See inner.

outer

outer(other: NDArray): NDArray

Outer product (flattens inputs, computes a[i] * b[j]). See outer.

tensordot

tensordot(other: NDArray, axes?: number | [number[], number[]]): NDArray | number | bigint | Complex

Tensor dot product along specified axes. Default contracts 2 axes. See tensordot.

trace

trace(): number | bigint | Complex

Sum of diagonal elements. See trace.

diagonal

diagonal(offset?: number, axis1?: number, axis2?: number): NDArray

Return specified diagonals. See diagonal.

Sorting and searching methods

sort

sort(axis?: number): NDArray

Return a sorted copy. Default sorts along the last axis. See sort.

argsort

argsort(axis?: number): NDArray

Return indices that would sort the array. See argsort.

partition

partition(kth: number, axis?: number): NDArray

Partially sort so that the element at kth position is in its final sorted position. See partition.

argpartition

argpartition(kth: number, axis?: number): NDArray

Return indices that would partition the array. See argpartition.

searchsorted

searchsorted(v: NDArray, side?: 'left' | 'right'): NDArray

Find indices where elements should be inserted to maintain order. See searchsorted.

nonzero

nonzero(): NDArray[]

Return indices of non-zero elements, one array per dimension. See nonzero.

argwhere

argwhere(): NDArray

Find indices of non-zero elements, returned as a 2D array of shape (N, ndim). See argwhere.

Indexing methods

take

take(indices: number[], axis?: number): NDArray

Take elements along an axis. See take.

put

put(indices: number[], values: NDArray | number | bigint): void

Set values at flat indices (in-place). See put.

choose

choose(choices: NDArray[]): NDArray

Construct an array by choosing from a list of arrays using this as the index. See choose.

compress

compress(condition: NDArray | boolean[], axis?: number): NDArray

Return selected slices along an axis where condition is true. See compress.

repeat

repeat(repeats: number | number[], axis?: number): NDArray

Repeat elements of the array. See repeat.

resize

resize(newShape: number[]): NDArray

Return a new array with the given shape. Repeats data if the new shape is larger. See resize.

bindex

bindex(mask: NDArray, axis?: number): NDArray

Boolean array indexing. Select elements where mask is true. See compress.

iindex

iindex(indices: NDArray | number[] | number[][], axis?: number): NDArray

Integer array indexing (fancy indexing). Select elements by an array of indices. See take.

Complex number methods

conj

conj(): NDArray

Return the complex conjugate, element-wise. For real arrays, returns a copy. See conj.

conjugate

conjugate(): NDArray

Alias for conj. See conj.

Other methods

copy

copy(): NDArray

Return a deep copy. See NDArray Class.

astype

astype(dtype: DType, copy?: boolean): NDArray

Cast to a new dtype. See NDArray Class.

fill

fill(value: number | bigint): void

Fill array in-place with a scalar. See NDArray Class.

toArray

toArray(): NestedArray

Convert to nested JavaScript array. See NDArray Class.

tolist

tolist(): NestedArray

Same as toArray. See NDArray Class.

tobytes

tobytes(): ArrayBuffer

Serialize to raw bytes. See NDArray Class.

get

get(indices: number[]): number | bigint | Complex

Get element by multi-dimensional indices. See NDArray Class.

set

set(indices: number[], value: number | bigint | Complex): void

Set element by multi-dimensional indices. See NDArray Class.

slice

slice(...sliceStrs: string[]): NDArray

NumPy-style string slicing. See NDArray Class.

row

row(i: number): NDArray

Get row i. See NDArray Class.

col

col(j: number): NDArray

Get column j. See NDArray Class.

rows

rows(start: number, stop: number): NDArray

Get a range of rows. See NDArray Class.

cols

cols(start: number, stop: number): NDArray

Get a range of columns. See NDArray Class.

item

item(...args: number[]): number | bigint | Complex

Extract a scalar value. See NDArray Class.

​NDArray Methods

​Arithmetic methods

​add

​subtract

​multiply

​divide

​floor_divide

​mod

​remainder

​fmod

​power

​negative

​positive

​absolute

​reciprocal

​square

​sqrt

​cbrt

​fabs

​sign

​heaviside

​divmod

​Exponential and logarithmic methods

​exp

​exp2

​expm1

​log

​log2

​log10

​log1p

​logaddexp

​logaddexp2

​Trigonometric methods

​sin

​cos

​tan

​arcsin

​arccos

​arctan

​arctan2

​hypot

​degrees

​radians

​Hyperbolic methods

​sinh

​cosh

​tanh

​arcsinh

​arccosh

​arctanh

​Rounding methods

​ceil

​floor

​fix

​rint

​trunc

​around

​round

​clip

​Comparison methods

​equal

​not_equal

​greater

​greater_equal

​less

​less_equal

​allclose

​isclose

​isnan

​isinf

​isfinite

​isnat

​Logic methods

​logical_and

​logical_or

​logical_not

​logical_xor

​invert

​bitwise_and

​bitwise_or

NDArray Methods

Arithmetic methods

add

subtract

multiply

divide

floor_divide

mod

remainder

fmod

power

negative

positive

absolute

reciprocal

square

sqrt

cbrt

fabs

sign

heaviside

divmod

Exponential and logarithmic methods

exp

exp2

expm1

log

log2

log10

log1p

logaddexp

logaddexp2

Trigonometric methods

sin

cos

tan

arcsin

arccos

arctan

arctan2

hypot

degrees

radians

Hyperbolic methods

sinh

cosh

tanh

arcsinh

arccosh

arctanh

Rounding methods

ceil

floor

fix

rint

trunc

around

round

clip

Comparison methods

equal

not_equal

greater

greater_equal

less

less_equal

allclose

isclose

isnan

isinf

isfinite

isnat

Logic methods

logical_and

logical_or

logical_not

logical_xor

invert

bitwise_and

bitwise_or