Function reference • torch

Tensor creation and casting Functions used to create new tensors and cast them back to R arrays.
`tensor()`	Creates a torch tensor.
`tensor_from_r()`	Create torch Tensor from R object
`as.array(<tensor>)`	Tensor casting
`as.matrix(<tensor>)`	Casts a 2d tensor to a matrix.
`tch_randn()`	Random normal
`tch_arange()`	arange
`tch_empty()`	Empty tensor
`tch_eye()`	Eye matrix (identity matrix)
`tch_full()`	Full tensor (one value filled tensor)
`tch_linspace()`	Linear spaced tensor
`tch_logspace()`	Loglinear spaced tensor
`tch_ones()`	One filled tensor
`tch_rand()`	Random uniform
`tch_randint()`	Random discrete uniform
`tch_randperm()`	Random permutation
`tch_zeros()`	Zero filled tensor
Autograd Utility functions related to autograd.
`with_no_grad()`	With no grad
Tensor methods Methods for tensors. They are always prefixed with `tch_`
`-`(<tensor>)	substraction
`==`(<tensor>)	Eq
`+`(<tensor>)	elementwise addition
`^`(<tensor>)	pow
`/`(<tensor>)	Div
`*`(<tensor>)	elementwise multiplciation
`tch_log()` `tch_log2()` `tch_log10()` `tch_log1p()`	log
`tch_abs()`	Abs
`tch_acos()`	Arc-cos
`tch_all()`	all
`tch_allclose()`	allclose
`tch_any()`	any
`tch_arange()`	arange
`tch_argmax()`	argmax
`tch_argmin()`	argmin
`tch_as_strided()`	as_strided
`tch_asin()`	asin
`tch_atan()`	atan
`tch_atan2()`	atan2
`tch_baddbmm()`	baddbmm
`tch_bernoulli()`	bernoulli
`tch_bincount()`	bincount
`tch_bmm()`	bmm
`tch_btrifact()`	Batch LU factorization
`tch_btrisolve()`	Batch LU solve
`tch_ceil()`	Ceil
`tch_chunk()`	Chunk
`tch_clamp()`	Clamp
`tch_clamp_max()`	Clamp max
`tch_clamp_min()`	Clamp min
`tch_cos()`	Cosine
`tch_cosh()`	Hyperbolic Cosine
`tch_cross()`	Cross
`tch_cuda_is_available()`	Checks if cuda is available.
`tch_cumprod()`	Cumprod
`tch_cumsum()`	Cumsum
`tch_det()`	Determinant
`tch_diag()`	Diag
`tch_diagflat()`	Diagflat
`tch_diagonal()`	Diagonal
`tch_digamma()`	Digamma
`tch_dist()`	Distance
`tch_dot()`	Dot product
`tch_eig()`	Eigenvalues and Eigenvectors
`tch_empty()`	Empty tensor
`tch_equal()`	Equal
`tch_eye()`	Eye matrix (identity matrix)
`tch_full()`	Full tensor (one value filled tensor)
`tch_gels()`	Gels
`tch_lerp()`	lerp
`tch_linspace()`	Linear spaced tensor
`tch_logspace()`	Loglinear spaced tensor
`tch_logsumexp()`	logsumexp
`tch_max()`	max
`tch_mean()`	mean
`tch_median()`	median
`tch_min()`	min
`tch_mm()`	matrix multiplication
`tch_mode()`	mode
`tch_ones()`	One filled tensor
`tch_permute()`	permute
`tch_prod()`	prod
`tch_qr()`	QR decomposition
`tch_rand()`	Random uniform
`tch_randint()`	Random discrete uniform
`tch_randn()`	Random normal
`tch_randperm()`	Random permutation
`tch_round()`	round
`tch_rsqrt()`	rsqrt
`tch_sigmoid()`	sigmoid
`tch_sign()`	sign
`tch_sin()`	sin
`tch_sinh()`	sinh
`tch_sqrt()`	sqrt
`tch_std()`	std
`tch_sum()`	sum
`tch_t()`	transpose
`tch_tan()`	tan
`tch_tanh()`	tanh
`tch_tril()`	Lower triangle
`tch_triu()`	Upper triangle
`tch_trunc()`	trunc
`tch_var()`	var
`tch_zeros()`	Zero filled tensor

Reference

Tensor creation and casting

Autograd

Tensor methods

Contents