clj-djl.ndarray

*

(* array0 array1)

*!

(*! array0 array1)

**

(** array0 array1)

+

(+ array0 array1)

+!

(+! array0 array1)

-

(- array0)(- array0 array1)

-!

(-! array0)(-! array0 array1)

substract element wise in place

/

(/ array0 array1)

<

(< array0 param1)

<=

(<= array0 param1)

=

(= array0 array1)

>

(> array0 param1)

>=

(>= array0 param1)

abs

(abs ndarray)

Returns the absolute value of this NDArray element-wise

acos

(acos ndarray)

Returns the inverse trigonometric cosine of this NDArray element-wise.

all-close

(all-close array0 array1)(all-close array0 array1 rtol atol equal-nan)

arange

multimethod

argmax

(argmax ndarray)(argmax ndarray axis)

argmin

(argmin ndarray)(argmin ndarray axis)

argsort

(argsort ndarray)(argsort ndarray axis)(argsort ndarray axis ascending)

base-manager

(base-manager)

concat

multimethod

copy

(copy ndarray)

create

multimethod

create-csr

(create-csr manager data indptr indices shape- & device)

create-row-sparse

(create-row-sparse manager data datashape indices shape- & device)

cumsum

(cumsum array)(cumsum array axis)

datatype

(datatype datatype-)

div

(div array0 array1)

dot

(dot array0 array1)

dup

duplicate

(duplicate ndarray)

exp

(exp array0)

expand-dims

(expand-dims ndarray axis)

eye

(eye manager rows)(eye manager rows k)(eye manager rows cols k)(eye manager rows cols k datatype-)(eye manager rows cols k datatype- device)

flatten

(flatten ndarray)

float-or-int

(float-or-int n)

full

(full manager shape- data)(full manager shape- data datatype-)(full manager shape- data datatype- device)

get

multimethod

get-device

(get-device ndarray)

get-element

(get-element array)(get-element array index)

get the wrap value of NDArray

get-gradient

(get-gradient ndarray)

Returns the gradient NDArray attached to this NDArray.

get-shape

(get-shape ndarray)

head

(head ndlist)

index

is-sparse

(is-sparse ndarray)

linspace

(linspace manager start stop n)(linspace manager start stop n endpoint)(linspace manager start stop n endpoint device)

log

(log ndarray)

log-softmax

(log-softmax ndarray axis)

log10

(log10 ndarray)

max

(max ndarray & [axes keep-dims])

mean

(mean ndarray & [axes keep-dims])

min

(min ndarray & [axes keep-dims])

mul

(mul array0 array1)

muli

(muli array0 array1)

ndindex

(ndindex)(ndindex param1 & more)

ndlist

multimethod

new-base-manager

new-ndindex

new-ndlist

multimethod

new-shape

norm

(norm ndarray)

ones

(ones manager shape-)(ones manager shape- datatype-)(ones manager shape- datatype- device)

ones-like

(ones-like ndarray)

pow

(pow array0 array1)

pp

(pp array)

prod

(prod ndarray & [axes keep-dims])

random-multinomial

(random-multinomial ndmanager n ndarray)(random-multinomial ndmanager n ndarray shape-)

Draw samples from a multinomial distribution.

random-normal

(random-normal manager shape-)(random-normal manager loc scale shape-)(random-normal manager loc scale shape- datatype-)(random-normal manager loc scale shape- datatype- device)

random-uniform

(random-uniform manager low high shape-)(random-uniform manager low high shape- datatype-)(random-uniform manager low high shape- datatype- device)

reshape

(reshape ndarray)(reshape ndarray param1 & more)

scalar?

(scalar? ndarray)

set

(set array data)(set array index val-or-fun)

set-requires-gradient

(set-requires-gradient ndarray requires-grad)

Attaches a gradient NDArray to this NDArray and marks it so GradientCollector.backward(NDArray) can compute the gradient with respect to it.

set-scalar

(set-scalar ndarray ndindex value)

shape

(shape)(shape param1 & more)

Create a shape instance.

singleton-or-throw

(singleton-or-throw ndlist)

size

(size ndarray)(size ndarray axis)

calc the seize of a ndarray.

sort

(sort ndarray)(sort ndarray axis)

sparse?

split

(split ndarray index-section & [axis])

sqrt

(sqrt ndarray)

squeeze

(squeeze ndarray & [axis])

stack

multimethod

sum

(sum array)(sum array axes)(sum array axes keep-dims)

t

to-array

(to-array ndarray-or-shape)

to-type

(to-type ndarray datatype- copy)

convert ndarray to data-type, available options are: “int8” “uint8” “int32” “int64” “float16” “float32” “float64” “boolean” “string” “unknown”

to-vec

(to-vec ndarray-or-shape)

trace

(trace ndarray offset axis1 axis2)(trace ndarray offset)(trace ndarray)

transpose

(transpose ndarray)(transpose ndarray axis & [more])

zeros

(zeros manager shape-)(zeros manager shape- datatype-)(zeros manager shape- datatype- device)

zeros-like

(zeros-like ndarray)