Package index • tensory

Package

tensory-package tensory: tensory: Tensory - Modern Tensor Operations for R

Tensor classes and constructors

The dense Tensor object, its matricized and decomposed relatives, and the constructors that build them.

tensor(): R6 Tensor Class
tenrand(): Random Dense Tensor
ones(): Create a tensor of ones
zeros(): Create a tensor of zeros
tendiag(): Diagonal Tensor
teneye(): Identity Tensor
tenfun(): Apply Elementwise Function to Tensor Arguments
tenmat(): R6 Tenmat Class
as.tenmat(): Convert object to Tenmat
ktensor(): R6 Class for Kruskal Tensors (KTensor)
ttensor(): R6 Class for Tucker Tensors (TTensor)
sptensor(): R6 Class for Sparse Tensors (Sptensor)
sptenmat(): R6 Class for Sparse Matricized Tensors (Sptenmat)
sptenrand(): Random Sparse Tensor
symtensor(): R6 Class for Symmetric Tensors (SymTensor)
symktensor(): R6 Class for Symmetric Kruskal Tensors (SymKTensor)
sumtensor(): R6 Class for Implicit Sums of Tensors (SumTensor)
as.tensor(): Convert object to Tensor
as.tensor(<KTensor>): S3 function to convert KTensor to full Tensor
as.tensor(<TTensor>): S3 function to convert TTensor to full Tensor
as.tensor(<Tenmat>): Convert Tenmat to Tensor

Shape, indexing, and predicates

Reshaping, unfolding, conversion, and structural queries.

permute(): Permute Tensor Dimensions
reshape(): Reshape Tensor
squeeze(): Squeeze Tensor
unfold(): Unfold Tensor
vec(): Vectorize Tensor
find(): Find Nonzero Entries
nnz(): Number of Nonzeros
full(): Dense Array Representation
isequal(): Equality Test for Tensors
isscalar(): Scalar Tensor Predicate
issymmetric(): Check Tensor Symmetry
symmetrize(): Symmetrize Tensor
transpose(): Transpose Tensor
double.Tensor(): MATLAB-Style Double Conversion
double.Tenmat(): Convert Tenmat to standard R double array (alias for matrix)
as.double(<Tenmat>): Convert Tenmat to standard R Matrix using generic type conversion
as.matrix(<Tenmat>): Convert Tenmat to standard R Matrix
as.vector(<Tenmat>): Convert Tenmat to standard R vector
head(<Tensor>): S3 head method for Tensor
tail(<Tensor>): S3 tail method for Tensor
show.Tensor(): S3 show method for Tensor
print(<Tensor>): S3 print method for Tensor
print(<KTensor>): S3 print method for KTensor
print(<TTensor>): S3 print method for TTensor
Math(<Tensor>): S3 Math group generic for Tensor
Summary(<Tensor>): S3 Summary group generic for Tensor

Products, contractions, and reductions

The computational core: tensor-times-matrix/vector/tensor products, matricized products, norms, and reductions.

ttm(): Tensor Times Matrix/Vector (ttm) Operation
ttv(): Tensor Times Vector
ttt(): Tensor Times Tensor (ttt) Operation
ttsv(): Tensor Times Same Vector
mtimes(): Matrix Multiplication Alias
`%*%`: S3 Matrix Multiplication Generic
innerprod(): Inner Product
contract(): Contract Tensor Dimensions
khatri_rao(): Khatri-Rao Product
kronecker(): Kronecker Product
hadamard(): Hadamard Product
mttkrp(): Matricized Tensor Times Khatri-Rao Product
mttkrps(): Sequence of MTTKRP Calculations
fibers(): Extract Tensor Fibers
mask(): Mask Tensor Values
collapse(): Collapse Tensor
scale(): Tensor Scaling
t_scale(): Scale Tensor
fnorm(): Frobenius Norm
nvecs(): Leading Mode-n Vectors

Decompositions

CP, Tucker, generalized CP, and tensor eigenpairs.

cp_als(): CP Alternating Least Squares Decomposition
cp_nmu(): Nonnegative CP Decomposition via Multiplicative Updates
cp_apr(): Poisson CP Decomposition (CP-APR) via Multiplicative Updates
cp_opt(): CP Decomposition via Direct Optimization
cp_wopt(): Weighted CP Decomposition via Direct Optimization
cp_arls(): CP Decomposition via Randomized (Sampled) ALS
cp_sym(): Symmetric CP Decomposition via Direct Optimization
gcp_opt(): Generalized CP Decomposition
tucker_als(): Tucker Alternating Least Squares (HOOI)
tucker_sym(): Symmetric Tucker Decomposition
hosvd(): Higher-Order Singular Value Decomposition
eig_sshopm(): Shifted Symmetric Higher-Order Power Method (SS-HOPM)
eig_geap(): Generalized Eigenproblem Adaptive Power Method (GEAP)

Regression with tensor predictors

Supervised models where each subject’s predictor is a whole array: sparse partial generalized tensor regression, tensor envelope PLS, and partial quantile tensor regression.

spgtr() coef(<spgtr>) print(<spgtr>): Sparse Partial Generalized Tensor Regression (SPGTR)
spgtr_cv(): Choose the Sparsity of a Tensor Regression by Cross-Validation
predict(<spgtr>): Predict from a Tensor Regression Fit
summary(<spgtr>): Summarize a Tensor Regression Fit
tepls(): Tensor Envelope Partial Least Squares Regression (TEPLS)
predict(<tepls>): Predict from a TEPLS Fit
pqtr() coef(<pqtr>) print(<pqtr>): Partial Quantile Tensor Regression (PQTR)
pqtr_cv(): Choose the Reduced Dimension of a Quantile Tensor Regression
predict(<pqtr>): Predict from a Partial Quantile Tensor Regression Fit

Working with CP factors

Post-processing, comparison, and visualization of KTensor fits.

arrange(): Arrange the Components of a Kruskal Tensor
normalize(): Normalize a Kruskal Tensor
fixsigns(): Fix Sign Ambiguity of a Kruskal Tensor
score(): Score the Similarity of Two Kruskal Tensors
ncomponents(): Number of Components of a Kruskal Tensor
extract(): Extract Components of a Kruskal Tensor
redistribute(): Redistribute Kruskal Weights into a Mode
tovec(): Kruskal Tensor to Vector
viz(): Visualize a Kruskal Tensor

Import and export

import_data(): Import Tensor Data from a Text File
export_data(): Export Tensor Data to a Text File