tests/testthat/prof_SuperBig.R
In rikenbit/DelayedTensor: R package for sparse and out-of-core arithmetic and decomposition of Tensor
# Setting
setVerbose(TRUE)
setAutoBlockSize(size=5E+8)
.Machine$integer.max = 10^12
load("../data/human_mid_brain.rda")
load("../data/mouse_mid_brain.rda")
darr_human_mini <- DelayedArray(as.array(human_mid_brain))
darr_mouse_mini <- DelayedArray(as.array(mouse_mid_brain))
# darr_human_mini <- DelayedArray(as.array(human_mid_brain[1:20, 1:20]))
# darr_mouse_mini <- DelayedArray(as.array(mouse_mid_brain[1:20, 1:20]))
#################################################################
# einsum
#################################################################
# total.time total.pct self.time self.pct
# "einsum" 0.16 100.0 0.00 0.0
# "initialize" 0.12 75.0 0.00 0.0
# "new" 0.12 75.0 0.00 0.0
# "validObject" 0.10 62.5 0.00 0.0
# "as" 0.08 50.0 0.00 0.0
# "asMethod" 0.08 50.0 0.00 0.0
# "DelayedArray" 0.08 50.0 0.00 0.0
# "HDF5ArraySeed" 0.08 50.0 0.00 0.0
# "new2" 0.08 50.0 0.00 0.0
print("einsum")
setSparse(FALSE)
Rprof()
darr <- einsum('ij,ik->ijk', darr_human_mini, darr_mouse_mini)
Rprof(NULL)
prof_einsum <- summaryRprof()
save(prof_einsum, file="prof_einsum.RData")
#################################################################
# vec (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".vec" 0.10 83.33 0.00 0.00
# "vec" 0.10 83.33 0.00 0.00
# "initialize" 0.06 50.00 0.02 16.67
# "new" 0.06 50.00 0.00 0.00
print("vec (Dense)")
setSparse(FALSE)
Rprof()
vec(darr)
Rprof(NULL)
prof_vec_dense <- summaryRprof()
save(prof_vec_dense, file="prof_vec_dense.RData")
#################################################################
# vec (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".vec" 0.10 100 0.00 0
# "vec" 0.10 100 0.00 0
# "as" 0.06 60 0.02 20
# "initialize" 0.06 60 0.00 0
# "new" 0.06 60 0.00 0
# "new2" 0.06 60 0.00 0
print("vec (Sparse)")
setSparse(TRUE)
Rprof()
v <- vec(darr)
Rprof(NULL)
prof_vec_sparse <- summaryRprof()
save(prof_vec_sparse, file="prof_vec_sparse.RData")
#################################################################
# unfold (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".unfold" 0.64 100.00 0.00 0.00
# "unfold" 0.64 100.00 0.00 0.00
# "as" 0.60 93.75 0.00 0.00
# ".class1" 0.58 90.62 0.00 0.00
# ".realize_and_return" 0.58 90.62 0.00 0.00
# ".vec" 0.58 90.62 0.00 0.00
# "realize" 0.58 90.62 0.00 0.00
# "vec" 0.58 90.62 0.00 0.00
# "[.data.frame" 0.52 81.25 0.00 0.00
# "[" 0.52 81.25 0.00 0.00
# "setAutoRealizationBackend" 0.52 81.25 0.00 0.00
# ".subset" 0.50 78.12 0.50 78.12
# ".get_realization_sink_constructor" 0.50 78.12 0.00 0.00
# "match" 0.50 78.12 0.00 0.00
print("unfold (Dense)")
setSparse(FALSE)
Rprof()
dmat <- unfold(darr, row_idx=1:2, col_idx=3)
Rprof(NULL)
prof_unfold_dense <- summaryRprof()
save(prof_unfold_dense, file="prof_unfold_dense.RData")
#################################################################
# unfold (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".unfold" 0.16 80 0.00 0
# "unfold" 0.16 80 0.00 0
# "as" 0.12 60 0.00 0
# ".class1" 0.10 50 0.00 0
# ".realize_and_return" 0.10 50 0.00 0
# "initialize" 0.10 50 0.00 0
# "new" 0.10 50 0.00 0
# "realize" 0.10 50 0.00 0
# "validObject" 0.10 50 0.00 0
print("unfold (Sparse)")
setSparse(TRUE)
Rprof()
unfold(darr, row_idx=1:2, col_idx=3)
Rprof(NULL)
prof_unfold_sparse <- summaryRprof()
save(prof_unfold_sparse, file="prof_unfold_sparse.RData")
#################################################################
# modeSum
#################################################################
# total.time total.pct self.time self.pct
# ".modeSum" 0.26 81.25 0.00 0.00
# "modeSum" 0.26 81.25 0.00 0.00
# "as" 0.22 68.75 0.00 0.00
# "asMethod" 0.22 68.75 0.00 0.00
# ".local" 0.18 56.25 0.00 0.00
# "initialize" 0.16 50.00 0.04 12.50
# ".class1" 0.16 50.00 0.00 0.00
# ".realize_and_return" 0.16 50.00 0.00 0.00
# "new" 0.16 50.00 0.00 0.00
# "new2" 0.16 50.00 0.00 0.00
# "realize" 0.16 50.00 0.00 0.00
# "writeHDF5Array" 0.16 50.00 0.00 0.00
print("modeSum")
setSparse(FALSE)
Rprof()
modeSum(darr, m=2)
Rprof(NULL)
prof_modesum <- summaryRprof()
save(prof_modesum, file="prof_modesum.RData")
#################################################################
# innerProd (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".BLOCK_Summary" 0.04 100 0.00 0
# ".innerProd" 0.04 100 0.00 0
# "blockReduce" 0.04 100 0.00 0
# "callNextMethod" 0.04 100 0.00 0
# "FUN" 0.04 100 0.00 0
# "innerProd" 0.04 100 0.00 0
# "sum" 0.04 100 0.00 0
print("innerProd (Dense)")
setSparse(FALSE)
Rprof()
innerProd(darr, darr)
Rprof(NULL)
prof_innerprod_dense <- summaryRprof()
save(prof_innerprod_dense, file="prof_innerprod_dense.RData")
#################################################################
# innerProd (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".Call2" 0.02 100 0.02 100
# ".BLOCK_Summary" 0.02 100 0.00 0
# ".h5mread2" 0.02 100 0.00 0
# ".innerProd" 0.02 100 0.00 0
# ".local" 0.02 100 0.00 0
# ".nextMethod" 0.02 100 0.00 0
# "blockReduce" 0.02 100 0.00 0
# "callNextMethod" 0.02 100 0.00 0
# "extract_array" 0.02 100 0.00 0
# "FUN" 0.02 100 0.00 0
# "gridReduce" 0.02 100 0.00 0
# "h5mread" 0.02 100 0.00 0
# "innerProd" 0.02 100 0.00 0
# "lapply" 0.02 100 0.00 0
# "read_block" 0.02 100 0.00 0
# "sum" 0.02 100 0.00 0
print("innerProd (Sparse)")
setSparse(TRUE)
Rprof()
innerProd(darr, darr)
Rprof(NULL)
prof_innerprod_sparse <- summaryRprof()
save(prof_innerprod_sparse, file="prof_innerprod_sparse.RData")
#################################################################
# hadamard (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".hadamard" 0.10 55.56 0.00 0.00
# "<Anonymous>" 0.10 55.56 0.00 0.00
# "hadamard" 0.10 55.56 0.00 0.00
# "new2" 0.08 44.44 0.02 11.11
# "show_compact_array" 0.08 44.44 0.00 0.00
print("hadamard (Dense)")
setSparse(FALSE)
Rprof()
hadamard(darr, darr)
Rprof(NULL)
prof_hadamard_dense <- summaryRprof()
save(prof_hadamard_dense, file="prof_hadamard_dense.RData")
#################################################################
# hadamard (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".hadamard" 0.24 80.00 0.00 0.00
# "hadamard" 0.24 80.00 0.00 0.00
# "FUN" 0.22 73.33 0.00 0.00
# "gridReduce" 0.22 73.33 0.00 0.00
# ".block_hadamard" 0.20 66.67 0.00 0.00
print("hadamard (Sparse)")
setSparse(TRUE)
Rprof()
hadamard(darr, darr)
Rprof(NULL)
prof_hadamard_sparse <- summaryRprof()
save(prof_hadamard_sparse, file="prof_hadamard_sparse.RData")
#################################################################
# kronecker (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".kronecker" 12.96 98.33 0.00 0.00
# "kronecker" 12.96 98.33 0.00 0.00
# "write_block" 7.70 58.42 0.00 0.00
# ".Call" 6.86 52.05 6.86 52.05
# "doTryCatch" 6.86 52.05 0.00 0.00
# "H5Dwrite" 6.86 52.05 0.00 0.00
# "h5write.default" 6.86 52.05 0.00 0.00
# "h5write" 6.86 52.05 0.00 0.00
# "h5writeDataset.array" 6.86 52.05 0.00 0.00
# "h5writeDataset" 6.86 52.05 0.00 0.00
# "h5writeDatasetHelper" 6.86 52.05 0.00 0.00
# "try" 6.86 52.05 0.00 0.00
# "tryCatch" 6.86 52.05 0.00 0.00
# "tryCatchList" 6.86 52.05 0.00 0.00
# "tryCatchOne" 6.86 52.05 0.00 0.00
print("kronecker (Dense)")
setSparse(FALSE)
Rprof()
kronecker(darr, darr)
Rprof(NULL)
prof_kronecker_dense <- summaryRprof()
save(prof_kronecker_dense, file="prof_kronecker_dense.RData")
#################################################################
# kronecker (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".kronecker" 10.76 97.82 0.00 0.00
# "kronecker" 10.76 97.82 0.00 0.00
# "write_block" 8.80 80.00 0.00 0.00
# "doTryCatch" 6.86 62.36 0.02 0.18
# "try" 6.86 62.36 0.00 0.00
# "tryCatch" 6.86 62.36 0.00 0.00
# "tryCatchList" 6.86 62.36 0.00 0.00
# "tryCatchOne" 6.86 62.36 0.00 0.00
# ".Call" 6.82 62.00 6.82 62.00
# "H5Dwrite" 6.82 62.00 0.00 0.00
# "h5write.default" 6.82 62.00 0.00 0.00
# "h5write" 6.82 62.00 0.00 0.00
# "h5writeDataset.array" 6.82 62.00 0.00 0.00
# "h5writeDataset" 6.82 62.00 0.00 0.00
# "h5writeDatasetHelper" 6.82 62.00 0.00 0.00
print("kronecker (Sparse)")
setSparse(TRUE)
Rprof()
kronecker(darr, darr)
Rprof(NULL)
prof_kronecker_sparse <- summaryRprof()
save(prof_kronecker_sparse, file="prof_kronecker_sparse.RData")
#################################################################
# khatri_rao (Dense)
#################################################################
# total.time total.pct self.time self.pct
# "khatri_rao" 0.14 87.5 0.00 0.0
# ".khatri_rao" 0.12 75.0 0.00 0.0
# "[[" 0.06 37.5 0.00 0.0
# "getArrayElement" 0.06 37.5 0.00 0.0
print("khatri_rao (Dense)")
setSparse(FALSE)
Rprof()
khatri_rao(darr[,,1], darr[,,1])
Rprof(NULL)
prof_khatri_rao_dense <- summaryRprof()
save(prof_khatri_rao_dense, file="prof_khatri_rao_dense.RData")
#################################################################
# khatri_rao (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".khatri_rao" 0.14 87.5 0.00 0
# "khatri_rao" 0.14 87.5 0.00 0
# "callNextMethod" 0.08 50.0 0.00 0
print("khatri_rao (Sparse)")
setSparse(TRUE)
Rprof()
khatri_rao(darr[,,1], darr[,,1])
Rprof(NULL)
prof_khatri_rao_sparse <- summaryRprof()
save(prof_khatri_rao_sparse, file="prof_khatri_rao_sparse.RData")
#################################################################
# fold (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".fold" 0.18 75.00 0.00 0.00
# ".reshapeIncNumbers1D" 0.18 75.00 0.00 0.00
# "fold" 0.18 75.00 0.00 0.00
# "new" 0.12 50.00 0.00 0.00
print("fold (Dense)")
setSparse(FALSE)
Rprof()
fold(dmat, row_idx=1:2, col_idx=3, dim(darr))
Rprof(NULL)
prof_fold_dense <- summaryRprof()
save(prof_fold_dense, file="prof_fold_dense.RData")
#################################################################
# fold (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".fold" 0.18 75.00 0.00 0.00
# ".reshapeIncNumbers1D" 0.18 75.00 0.00 0.00
# "fold" 0.18 75.00 0.00 0.00
# "initialize" 0.14 58.33 0.04 16.67
# "new" 0.14 58.33 0.00 0.00
print("fold (Sparse)")
setSparse(TRUE)
Rprof()
fold(dmat, row_idx=1:2, col_idx=3, dim(darr))
Rprof(NULL)
prof_fold_sparse <- summaryRprof()
save(prof_fold_sparse, file="prof_fold_sparse.RData")
#################################################################
# diag
#################################################################
# total.time total.pct self.time self.pct
# ".diag" 0.18 90 0.00 0
# "[" 0.18 90 0.00 0
# "DelayedTensor::diag" 0.18 90 0.00 0
# "eval" 0.18 90 0.00 0
# "drop" 0.10 50 0.00 0
# "standardGeneric" 0.10 50 0.00 0
print("diag")
Rprof()
DelayedTensor::diag(darr)
Rprof(NULL)
prof_diag1 <- summaryRprof()
save(prof_diag1, file="prof_diag1.RData")
#################################################################
# "diag<-"
#################################################################
# total.time total.pct self.time self.pct
# "DelayedTensor::diag<-" 0.26 100.00 0.00 0.00
# "[<-" 0.18 69.23 0.00 0.00
# "eval" 0.18 69.23 0.00 0.00
print("diag<-")
Rprof()
DelayedTensor::diag(darr) <- seq(min(dim(darr)))
Rprof(NULL)
prof_diag2 <- summaryRprof()
save(prof_diag2, file="prof_diag2.RData")
#################################################################
# modebind_list (Dense)
#################################################################
# total.time total.pct self.time self.pct
# "modebind_list" 0.34 85 0.00 0
# "new2" 0.22 55 0.02 5
# "FUN" 0.22 55 0.00 0
# "initialize" 0.22 55 0.00 0
# "new" 0.22 55 0.00 0
# "validObject" 0.22 55 0.00 0
print("modebind_list (Dense)")
setSparse(FALSE)
Rprof()
modebind_list(list(darr, darr), m=1)
Rprof(NULL)
prof_modebind_list_dense <- summaryRprof()
save(prof_modebind_list_dense, file="prof_modebind_list_dense.RData")
#################################################################
# modebind_list (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# "modebind_list" 0.32 84.21 0.00 0.00
# "FUN" 0.24 63.16 0.00 0.00
# "lapply" 0.22 57.89 0.00 0.00
# "initialize" 0.20 52.63 0.04 10.53
# "as" 0.20 52.63 0.00 0.00
# "asMethod" 0.20 52.63 0.00 0.00
# "new" 0.20 52.63 0.00 0.00
print("modebind_list (Sparse)")
setSparse(TRUE)
Rprof()
modebind_list(list(darr, darr), m=1)
Rprof(NULL)
prof_modebind_list_sparse <- summaryRprof()
save(prof_modebind_list_sparse, file="prof_modebind_list_sparse.RData")
#################################################################
# DelayedDiagonalArray
#################################################################
# total.time total.pct self.time self.pct
# ".print_2D_slices" 0.04 100 0.00 0
# ".print_array_data" 0.04 100 0.00 0
# ".print_nDarray_data" 0.04 100 0.00 0
# "<Anonymous>" 0.04 100 0.00 0
# "show_compact_array" 0.04 100 0.00 0
print("DelayedDiagonalArray")
Rprof()
DelayedDiagonalArray(c(100,200,300), 1:100)
Rprof(NULL)
prof_delayeddiagonalarray <- summaryRprof()
save(prof_delayeddiagonalarray, file="prof_delayeddiagonalarray.RData")
#################################################################
# hosvd (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".hosvd" 10.92 98.73 0.00 0.00
# "hosvd" 10.92 98.73 0.00 0.00
# "DelayedArray" 9.66 87.34 0.00 0.00
# "<Anonymous>" 9.04 81.74 0.00 0.00
# "do.call" 8.90 80.47 0.02 0.18
# "suppressWarnings" 8.62 77.94 0.02 0.18
# "withCallingHandlers" 8.62 77.94 0.02 0.18
# ".svd" 8.62 77.94 0.00 0.00
# "runIrlbaSVD" 8.62 77.94 0.00 0.00
# ".super_BLOCK_mult" 8.60 77.76 0.00 0.00
# "mult" 8.22 74.32 0.00 0.00
# "as" 8.00 72.33 0.00 0.00
# "standardGeneric" 7.82 70.71 0.12 1.08
# "realize" 7.66 69.26 0.00 0.00
# "drop" 7.56 68.35 0.00 0.00
# "asMethod" 7.40 66.91 0.00 0.00
# ".local" 7.28 65.82 0.00 0.00
# "writeHDF5Array" 6.62 59.86 0.00 0.00
print("hosvd (Dense)")
setSparse(FALSE)
Rprof()
hosvd(darr, ranks=c(2,3,4))
Rprof(NULL)
prof_hosvd_dense <- summaryRprof()
save(prof_hosvd_dense, file="prof_hosvd_dense.RData")
#################################################################
# hosvd (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".hosvd" 2.28 94.21 0.00 0.00
# "hosvd" 2.28 94.21 0.00 0.00
# "ttl" 1.88 77.69 0.00 0.00
# "as" 1.68 69.42 0.00 0.00
# "new" 1.38 57.02 0.06 2.48
# "initialize" 1.34 55.37 0.18 7.44
# "asMethod" 1.26 52.07 0.00 0.00
print("hosvd (Sparse)")
setSparse(TRUE)
Rprof()
hosvd(darr, ranks=c(2,3,4))
Rprof(NULL)
prof_hosvd_sparse <- summaryRprof()
save(prof_hosvd_sparse, file="prof_hosvd_sparse.RData")
#################################################################
# cp (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".cp" 5.52 97.87 0.00 0.00
# "cp" 5.52 97.87 0.00 0.00
# "as" 4.04 71.63 0.02 0.35
# "ttl" 3.24 57.45 0.00 0.00
# "new" 3.00 53.19 0.02 0.35
# "initialize" 2.96 52.48 0.42 7.45
print("cp (Dense)")
setSparse(FALSE)
Rprof()
cp(darr, num_components=2, max_iter=2)
Rprof(NULL)
prof_cp_dense <- summaryRprof()
save(prof_cp_dense, file="prof_cp_dense.RData")
#################################################################
# cp (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".cp" 4.76 97.94 0.00 0.00
# "cp" 4.76 97.94 0.00 0.00
# "as" 3.28 67.49 0.00 0.00
# "ttl" 2.86 58.85 0.00 0.00
setSparse(TRUE)
Rprof()
cp(darr, num_components=2, max_iter=2)
Rprof(NULL)
prof_cp_sparse <- summaryRprof()
save(prof_cp_sparse, file="prof_cp_sparse.RData")
#################################################################
# tucker (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".tucker" 12.82 99.07 0.00 0.00
# "tucker" 12.82 99.07 0.00 0.00
# "DelayedArray" 10.20 78.83 0.02 0.15
# "as" 9.36 72.33 0.02 0.15
# "<Anonymous>" 8.54 66.00 0.04 0.31
# ".super_BLOCK_mult" 8.44 65.22 0.00 0.00
# "do.call" 8.40 64.91 0.00 0.00
# ".svd" 8.34 64.45 0.00 0.00
# "realize" 8.32 64.30 0.00 0.00
# "asMethod" 8.26 63.83 0.00 0.00
# "withCallingHandlers" 8.02 61.98 0.02 0.15
# "suppressWarnings" 8.02 61.98 0.00 0.00
# "runIrlbaSVD" 8.00 61.82 0.00 0.00
# "standardGeneric" 7.88 60.90 0.12 0.93
# "mult" 7.68 59.35 0.00 0.00
# ".local" 7.42 57.34 0.00 0.00
# "drop" 7.38 57.03 0.00 0.00
# "writeHDF5Array" 6.78 52.40 0.00 0.00
print("tucker (Dense)")
setSparse(FALSE)
Rprof()
tucker(darr, ranks=c(2,2,2), max_iter=2)
Rprof(NULL)
prof_tucker_dense <- summaryRprof()
save(prof_tucker_dense, file="prof_tucker_dense.RData")
#################################################################
# tucker (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".tucker" 4.84 97.58 0.00 0.00
# "tucker" 4.84 97.58 0.00 0.00
# "ttl" 3.74 75.40 0.00 0.00
# "as" 3.58 72.18 0.04 0.81
# "asMethod" 2.60 52.42 0.00 0.00
# "new" 2.56 51.61 0.12 2.42
# "initialize" 2.50 50.40 0.32 6.45
print("tucker (Sparse)")
setSparse(TRUE)
Rprof()
tucker(darr, ranks=c(2,2,2), max_iter=2)
Rprof(NULL)
prof_tucker_sparse <- summaryRprof()
save(prof_tucker_sparse, file="prof_tucker_sparse.RData")
#################################################################
# mpca (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".mpca" 13.64 98.70 0.00 0.00
# "mpca" 13.64 98.70 0.00 0.00
# "DelayedArray" 11.40 82.49 0.00 0.00
# "<Anonymous>" 10.72 77.57 0.02 0.14
# "do.call" 10.54 76.27 0.00 0.00
# ".super_BLOCK_mult" 10.14 73.37 0.02 0.14
# ".svd" 10.04 72.65 0.00 0.00
# "as" 9.96 72.07 0.06 0.43
# "suppressWarnings" 9.82 71.06 0.02 0.14
# "runIrlbaSVD" 9.82 71.06 0.00 0.00
# "withCallingHandlers" 9.82 71.06 0.00 0.00
# "standardGeneric" 9.48 68.60 0.18 1.30
# "realize" 9.44 68.31 0.00 0.00
# "mult" 9.40 68.02 0.02 0.14
# "asMethod" 9.28 67.15 0.02 0.14
# "drop" 9.10 65.85 0.00 0.00
# ".local" 8.84 63.97 0.04 0.29
# "writeHDF5Array" 8.20 59.33 0.00 0.00
print("mpca (Dense)")
setSparse(FALSE)
Rprof()
mpca(darr, ranks=c(2,2), max_iter=2)
Rprof(NULL)
prof_mpca_dense <- summaryRprof()
save(prof_mpca_dense, file="prof_mpca_dense.RData")
#################################################################
# mpca (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".mpca" 3.70 95.36 0.00 0.00
# "mpca" 3.70 95.36 0.00 0.00
# "as" 2.60 67.01 0.00 0.00
# "ttl" 2.58 66.49 0.00 0.00
# "new" 2.12 54.64 0.00 0.00
# "initialize" 2.10 54.12 0.42 10.82
# "asMethod" 1.86 47.94 0.00 0.00
# ".reshapeIncNumbers1D" 1.74 44.85 0.00 0.00
# "DelayedArray" 1.74 44.85 0.00 0.00
# "realize" 1.74 44.85 0.00 0.00
print("mpca (Sparse)")
setSparse(TRUE)
Rprof()
mpca(darr, ranks=c(2,2), max_iter=2)
Rprof(NULL)
prof_mpca_sparse <- summaryRprof()
save(prof_mpca_sparse, file="prof_mpca_sparse.RData")
#################################################################
# pvd (Dense)
#################################################################
print("pvd (Dense)")
setSparse(FALSE)
Rprof()
pvd(darr, uranks=rep(2, dim(darr)[3]), wranks=rep(3, dim(darr)[3]), a=2, b=3)
Rprof(NULL)
prof_pvd_dense <- summaryRprof()
save(prof_pvd_dense, file="prof_pvd_dense.RData")
#################################################################
# pvd (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".pvd" 15.64 98.24 0.00 0.00
# "pvd" 15.64 98.24 0.00 0.00
# "FUN" 13.30 83.54 0.08 0.50
# "lapply" 13.28 83.42 0.08 0.50
# "as" 13.10 82.29 0.12 0.75
# ".super_BLOCK_mult" 9.76 61.31 0.00 0.00
# "%*%" 9.76 61.31 0.00 0.00
# ".class1" 9.66 60.68 0.00 0.00
# "realize" 9.50 59.67 0.00 0.00
# "asMethod" 9.30 58.42 0.04 0.25
# "bplapply2" 9.16 57.54 0.00 0.00
print("pvd (Sparse)")
setSparse(TRUE)
Rprof()
pvd(darr, uranks=rep(2, dim(darr)[3]), wranks=rep(3, dim(darr)[3]), a=2, b=3)
Rprof(NULL)
prof_pvd_sparse <- summaryRprof()
save(prof_pvd_sparse, file="prof_pvd_sparse.RData")
# Sparse mode is effective in the most cases
# > prof_vec_dense$sampling.time
# [1] 0.12
# > prof_vec_sparse$sampling.time
# [1] 0.1
# > prof_unfold_dense$sampling.time
# [1] 0.64
# > prof_unfold_sparse$sampling.time
# [1] 0.2
# > prof_innerprod_dense$sampling.time
# [1] 0.04
# > prof_innerprod_sparse$sampling.time
# [1] 0.02
# > prof_hadamard_dense$sampling.time
# [1] 0.18
# > prof_hadamard_sparse$sampling.time
# [1] 0.3
# > prof_kronecker_dense$sampling.time
# [1] 13.18
# > prof_kronecker_sparse$sampling.time
# [1] 11
# > prof_khatri_rao_dense$sampling.time
# [1] 0.16
# > prof_khatri_rao_sparse$sampling.time
# [1] 0.16
# > prof_fold_dense$sampling.time
# [1] 0.24
# > prof_fold_sparse$sampling.time
# [1] 0.24
# > prof_modebind_list_dense$sampling.time
# [1] 0.4
# > prof_modebind_list_sparse$sampling.time
# [1] 0.38
# > prof_hosvd_dense$sampling.time
# [1] 11.06
# > prof_hosvd_sparse$sampling.time
# [1] 2.42
# > prof_cp_dense$sampling.time
# [1] 5.64
# > prof_cp_sparse$sampling.time
# [1] 4.86
# > prof_tucker_dense$sampling.time
# [1] 12.94
# > prof_tucker_sparse$sampling.time
# [1] 4.96
# > prof_mpca_dense$sampling.time
# [1] 13.82
# > prof_mpca_sparse$sampling.time
# [1] 3.88
# > prof_pvd_dense$sampling.time
# [1] 0
# > prof_pvd_sparse$sampling.time
# [1] 15.92
rikenbit/DelayedTensor documentation built on Jan. 30, 2023, 6:15 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
# Setting
setVerbose(TRUE)
setAutoBlockSize(size=5E+8)
.Machine$integer.max = 10^12
load("../data/human_mid_brain.rda")
load("../data/mouse_mid_brain.rda")
darr_human_mini <- DelayedArray(as.array(human_mid_brain))
darr_mouse_mini <- DelayedArray(as.array(mouse_mid_brain))
# darr_human_mini <- DelayedArray(as.array(human_mid_brain[1:20, 1:20]))
# darr_mouse_mini <- DelayedArray(as.array(mouse_mid_brain[1:20, 1:20]))
#################################################################
# einsum
#################################################################
# total.time total.pct self.time self.pct
# "einsum" 0.16 100.0 0.00 0.0
# "initialize" 0.12 75.0 0.00 0.0
# "new" 0.12 75.0 0.00 0.0
# "validObject" 0.10 62.5 0.00 0.0
# "as" 0.08 50.0 0.00 0.0
# "asMethod" 0.08 50.0 0.00 0.0
# "DelayedArray" 0.08 50.0 0.00 0.0
# "HDF5ArraySeed" 0.08 50.0 0.00 0.0
# "new2" 0.08 50.0 0.00 0.0
print("einsum")
setSparse(FALSE)
Rprof()
darr <- einsum('ij,ik->ijk', darr_human_mini, darr_mouse_mini)
Rprof(NULL)
prof_einsum <- summaryRprof()
save(prof_einsum, file="prof_einsum.RData")
#################################################################
# vec (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".vec" 0.10 83.33 0.00 0.00
# "vec" 0.10 83.33 0.00 0.00
# "initialize" 0.06 50.00 0.02 16.67
# "new" 0.06 50.00 0.00 0.00
print("vec (Dense)")
setSparse(FALSE)
Rprof()
vec(darr)
Rprof(NULL)
prof_vec_dense <- summaryRprof()
save(prof_vec_dense, file="prof_vec_dense.RData")
#################################################################
# vec (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".vec" 0.10 100 0.00 0
# "vec" 0.10 100 0.00 0
# "as" 0.06 60 0.02 20
# "initialize" 0.06 60 0.00 0
# "new" 0.06 60 0.00 0
# "new2" 0.06 60 0.00 0
print("vec (Sparse)")
setSparse(TRUE)
Rprof()
v <- vec(darr)
Rprof(NULL)
prof_vec_sparse <- summaryRprof()
save(prof_vec_sparse, file="prof_vec_sparse.RData")
#################################################################
# unfold (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".unfold" 0.64 100.00 0.00 0.00
# "unfold" 0.64 100.00 0.00 0.00
# "as" 0.60 93.75 0.00 0.00
# ".class1" 0.58 90.62 0.00 0.00
# ".realize_and_return" 0.58 90.62 0.00 0.00
# ".vec" 0.58 90.62 0.00 0.00
# "realize" 0.58 90.62 0.00 0.00
# "vec" 0.58 90.62 0.00 0.00
# "[.data.frame" 0.52 81.25 0.00 0.00
# "[" 0.52 81.25 0.00 0.00
# "setAutoRealizationBackend" 0.52 81.25 0.00 0.00
# ".subset" 0.50 78.12 0.50 78.12
# ".get_realization_sink_constructor" 0.50 78.12 0.00 0.00
# "match" 0.50 78.12 0.00 0.00
print("unfold (Dense)")
setSparse(FALSE)
Rprof()
dmat <- unfold(darr, row_idx=1:2, col_idx=3)
Rprof(NULL)
prof_unfold_dense <- summaryRprof()
save(prof_unfold_dense, file="prof_unfold_dense.RData")
#################################################################
# unfold (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".unfold" 0.16 80 0.00 0
# "unfold" 0.16 80 0.00 0
# "as" 0.12 60 0.00 0
# ".class1" 0.10 50 0.00 0
# ".realize_and_return" 0.10 50 0.00 0
# "initialize" 0.10 50 0.00 0
# "new" 0.10 50 0.00 0
# "realize" 0.10 50 0.00 0
# "validObject" 0.10 50 0.00 0
print("unfold (Sparse)")
setSparse(TRUE)
Rprof()
unfold(darr, row_idx=1:2, col_idx=3)
Rprof(NULL)
prof_unfold_sparse <- summaryRprof()
save(prof_unfold_sparse, file="prof_unfold_sparse.RData")
#################################################################
# modeSum
#################################################################
# total.time total.pct self.time self.pct
# ".modeSum" 0.26 81.25 0.00 0.00
# "modeSum" 0.26 81.25 0.00 0.00
# "as" 0.22 68.75 0.00 0.00
# "asMethod" 0.22 68.75 0.00 0.00
# ".local" 0.18 56.25 0.00 0.00
# "initialize" 0.16 50.00 0.04 12.50
# ".class1" 0.16 50.00 0.00 0.00
# ".realize_and_return" 0.16 50.00 0.00 0.00
# "new" 0.16 50.00 0.00 0.00
# "new2" 0.16 50.00 0.00 0.00
# "realize" 0.16 50.00 0.00 0.00
# "writeHDF5Array" 0.16 50.00 0.00 0.00
print("modeSum")
setSparse(FALSE)
Rprof()
modeSum(darr, m=2)
Rprof(NULL)
prof_modesum <- summaryRprof()
save(prof_modesum, file="prof_modesum.RData")
#################################################################
# innerProd (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".BLOCK_Summary" 0.04 100 0.00 0
# ".innerProd" 0.04 100 0.00 0
# "blockReduce" 0.04 100 0.00 0
# "callNextMethod" 0.04 100 0.00 0
# "FUN" 0.04 100 0.00 0
# "innerProd" 0.04 100 0.00 0
# "sum" 0.04 100 0.00 0
print("innerProd (Dense)")
setSparse(FALSE)
Rprof()
innerProd(darr, darr)
Rprof(NULL)
prof_innerprod_dense <- summaryRprof()
save(prof_innerprod_dense, file="prof_innerprod_dense.RData")
#################################################################
# innerProd (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".Call2" 0.02 100 0.02 100
# ".BLOCK_Summary" 0.02 100 0.00 0
# ".h5mread2" 0.02 100 0.00 0
# ".innerProd" 0.02 100 0.00 0
# ".local" 0.02 100 0.00 0
# ".nextMethod" 0.02 100 0.00 0
# "blockReduce" 0.02 100 0.00 0
# "callNextMethod" 0.02 100 0.00 0
# "extract_array" 0.02 100 0.00 0
# "FUN" 0.02 100 0.00 0
# "gridReduce" 0.02 100 0.00 0
# "h5mread" 0.02 100 0.00 0
# "innerProd" 0.02 100 0.00 0
# "lapply" 0.02 100 0.00 0
# "read_block" 0.02 100 0.00 0
# "sum" 0.02 100 0.00 0
print("innerProd (Sparse)")
setSparse(TRUE)
Rprof()
innerProd(darr, darr)
Rprof(NULL)
prof_innerprod_sparse <- summaryRprof()
save(prof_innerprod_sparse, file="prof_innerprod_sparse.RData")
#################################################################
# hadamard (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".hadamard" 0.10 55.56 0.00 0.00
# "<Anonymous>" 0.10 55.56 0.00 0.00
# "hadamard" 0.10 55.56 0.00 0.00
# "new2" 0.08 44.44 0.02 11.11
# "show_compact_array" 0.08 44.44 0.00 0.00
print("hadamard (Dense)")
setSparse(FALSE)
Rprof()
hadamard(darr, darr)
Rprof(NULL)
prof_hadamard_dense <- summaryRprof()
save(prof_hadamard_dense, file="prof_hadamard_dense.RData")
#################################################################
# hadamard (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".hadamard" 0.24 80.00 0.00 0.00
# "hadamard" 0.24 80.00 0.00 0.00
# "FUN" 0.22 73.33 0.00 0.00
# "gridReduce" 0.22 73.33 0.00 0.00
# ".block_hadamard" 0.20 66.67 0.00 0.00
print("hadamard (Sparse)")
setSparse(TRUE)
Rprof()
hadamard(darr, darr)
Rprof(NULL)
prof_hadamard_sparse <- summaryRprof()
save(prof_hadamard_sparse, file="prof_hadamard_sparse.RData")
#################################################################
# kronecker (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".kronecker" 12.96 98.33 0.00 0.00
# "kronecker" 12.96 98.33 0.00 0.00
# "write_block" 7.70 58.42 0.00 0.00
# ".Call" 6.86 52.05 6.86 52.05
# "doTryCatch" 6.86 52.05 0.00 0.00
# "H5Dwrite" 6.86 52.05 0.00 0.00
# "h5write.default" 6.86 52.05 0.00 0.00
# "h5write" 6.86 52.05 0.00 0.00
# "h5writeDataset.array" 6.86 52.05 0.00 0.00
# "h5writeDataset" 6.86 52.05 0.00 0.00
# "h5writeDatasetHelper" 6.86 52.05 0.00 0.00
# "try" 6.86 52.05 0.00 0.00
# "tryCatch" 6.86 52.05 0.00 0.00
# "tryCatchList" 6.86 52.05 0.00 0.00
# "tryCatchOne" 6.86 52.05 0.00 0.00
print("kronecker (Dense)")
setSparse(FALSE)
Rprof()
kronecker(darr, darr)
Rprof(NULL)
prof_kronecker_dense <- summaryRprof()
save(prof_kronecker_dense, file="prof_kronecker_dense.RData")
#################################################################
# kronecker (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".kronecker" 10.76 97.82 0.00 0.00
# "kronecker" 10.76 97.82 0.00 0.00
# "write_block" 8.80 80.00 0.00 0.00
# "doTryCatch" 6.86 62.36 0.02 0.18
# "try" 6.86 62.36 0.00 0.00
# "tryCatch" 6.86 62.36 0.00 0.00
# "tryCatchList" 6.86 62.36 0.00 0.00
# "tryCatchOne" 6.86 62.36 0.00 0.00
# ".Call" 6.82 62.00 6.82 62.00
# "H5Dwrite" 6.82 62.00 0.00 0.00
# "h5write.default" 6.82 62.00 0.00 0.00
# "h5write" 6.82 62.00 0.00 0.00
# "h5writeDataset.array" 6.82 62.00 0.00 0.00
# "h5writeDataset" 6.82 62.00 0.00 0.00
# "h5writeDatasetHelper" 6.82 62.00 0.00 0.00
print("kronecker (Sparse)")
setSparse(TRUE)
Rprof()
kronecker(darr, darr)
Rprof(NULL)
prof_kronecker_sparse <- summaryRprof()
save(prof_kronecker_sparse, file="prof_kronecker_sparse.RData")
#################################################################
# khatri_rao (Dense)
#################################################################
# total.time total.pct self.time self.pct
# "khatri_rao" 0.14 87.5 0.00 0.0
# ".khatri_rao" 0.12 75.0 0.00 0.0
# "[[" 0.06 37.5 0.00 0.0
# "getArrayElement" 0.06 37.5 0.00 0.0
print("khatri_rao (Dense)")
setSparse(FALSE)
Rprof()
khatri_rao(darr[,,1], darr[,,1])
Rprof(NULL)
prof_khatri_rao_dense <- summaryRprof()
save(prof_khatri_rao_dense, file="prof_khatri_rao_dense.RData")
#################################################################
# khatri_rao (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".khatri_rao" 0.14 87.5 0.00 0
# "khatri_rao" 0.14 87.5 0.00 0
# "callNextMethod" 0.08 50.0 0.00 0
print("khatri_rao (Sparse)")
setSparse(TRUE)
Rprof()
khatri_rao(darr[,,1], darr[,,1])
Rprof(NULL)
prof_khatri_rao_sparse <- summaryRprof()
save(prof_khatri_rao_sparse, file="prof_khatri_rao_sparse.RData")
#################################################################
# fold (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".fold" 0.18 75.00 0.00 0.00
# ".reshapeIncNumbers1D" 0.18 75.00 0.00 0.00
# "fold" 0.18 75.00 0.00 0.00
# "new" 0.12 50.00 0.00 0.00
print("fold (Dense)")
setSparse(FALSE)
Rprof()
fold(dmat, row_idx=1:2, col_idx=3, dim(darr))
Rprof(NULL)
prof_fold_dense <- summaryRprof()
save(prof_fold_dense, file="prof_fold_dense.RData")
#################################################################
# fold (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".fold" 0.18 75.00 0.00 0.00
# ".reshapeIncNumbers1D" 0.18 75.00 0.00 0.00
# "fold" 0.18 75.00 0.00 0.00
# "initialize" 0.14 58.33 0.04 16.67
# "new" 0.14 58.33 0.00 0.00
print("fold (Sparse)")
setSparse(TRUE)
Rprof()
fold(dmat, row_idx=1:2, col_idx=3, dim(darr))
Rprof(NULL)
prof_fold_sparse <- summaryRprof()
save(prof_fold_sparse, file="prof_fold_sparse.RData")
#################################################################
# diag
#################################################################
# total.time total.pct self.time self.pct
# ".diag" 0.18 90 0.00 0
# "[" 0.18 90 0.00 0
# "DelayedTensor::diag" 0.18 90 0.00 0
# "eval" 0.18 90 0.00 0
# "drop" 0.10 50 0.00 0
# "standardGeneric" 0.10 50 0.00 0
print("diag")
Rprof()
DelayedTensor::diag(darr)
Rprof(NULL)
prof_diag1 <- summaryRprof()
save(prof_diag1, file="prof_diag1.RData")
#################################################################
# "diag<-"
#################################################################
# total.time total.pct self.time self.pct
# "DelayedTensor::diag<-" 0.26 100.00 0.00 0.00
# "[<-" 0.18 69.23 0.00 0.00
# "eval" 0.18 69.23 0.00 0.00
print("diag<-")
Rprof()
DelayedTensor::diag(darr) <- seq(min(dim(darr)))
Rprof(NULL)
prof_diag2 <- summaryRprof()
save(prof_diag2, file="prof_diag2.RData")
#################################################################
# modebind_list (Dense)
#################################################################
# total.time total.pct self.time self.pct
# "modebind_list" 0.34 85 0.00 0
# "new2" 0.22 55 0.02 5
# "FUN" 0.22 55 0.00 0
# "initialize" 0.22 55 0.00 0
# "new" 0.22 55 0.00 0
# "validObject" 0.22 55 0.00 0
print("modebind_list (Dense)")
setSparse(FALSE)
Rprof()
modebind_list(list(darr, darr), m=1)
Rprof(NULL)
prof_modebind_list_dense <- summaryRprof()
save(prof_modebind_list_dense, file="prof_modebind_list_dense.RData")
#################################################################
# modebind_list (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# "modebind_list" 0.32 84.21 0.00 0.00
# "FUN" 0.24 63.16 0.00 0.00
# "lapply" 0.22 57.89 0.00 0.00
# "initialize" 0.20 52.63 0.04 10.53
# "as" 0.20 52.63 0.00 0.00
# "asMethod" 0.20 52.63 0.00 0.00
# "new" 0.20 52.63 0.00 0.00
print("modebind_list (Sparse)")
setSparse(TRUE)
Rprof()
modebind_list(list(darr, darr), m=1)
Rprof(NULL)
prof_modebind_list_sparse <- summaryRprof()
save(prof_modebind_list_sparse, file="prof_modebind_list_sparse.RData")
#################################################################
# DelayedDiagonalArray
#################################################################
# total.time total.pct self.time self.pct
# ".print_2D_slices" 0.04 100 0.00 0
# ".print_array_data" 0.04 100 0.00 0
# ".print_nDarray_data" 0.04 100 0.00 0
# "<Anonymous>" 0.04 100 0.00 0
# "show_compact_array" 0.04 100 0.00 0
print("DelayedDiagonalArray")
Rprof()
DelayedDiagonalArray(c(100,200,300), 1:100)
Rprof(NULL)
prof_delayeddiagonalarray <- summaryRprof()
save(prof_delayeddiagonalarray, file="prof_delayeddiagonalarray.RData")
#################################################################
# hosvd (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".hosvd" 10.92 98.73 0.00 0.00
# "hosvd" 10.92 98.73 0.00 0.00
# "DelayedArray" 9.66 87.34 0.00 0.00
# "<Anonymous>" 9.04 81.74 0.00 0.00
# "do.call" 8.90 80.47 0.02 0.18
# "suppressWarnings" 8.62 77.94 0.02 0.18
# "withCallingHandlers" 8.62 77.94 0.02 0.18
# ".svd" 8.62 77.94 0.00 0.00
# "runIrlbaSVD" 8.62 77.94 0.00 0.00
# ".super_BLOCK_mult" 8.60 77.76 0.00 0.00
# "mult" 8.22 74.32 0.00 0.00
# "as" 8.00 72.33 0.00 0.00
# "standardGeneric" 7.82 70.71 0.12 1.08
# "realize" 7.66 69.26 0.00 0.00
# "drop" 7.56 68.35 0.00 0.00
# "asMethod" 7.40 66.91 0.00 0.00
# ".local" 7.28 65.82 0.00 0.00
# "writeHDF5Array" 6.62 59.86 0.00 0.00
print("hosvd (Dense)")
setSparse(FALSE)
Rprof()
hosvd(darr, ranks=c(2,3,4))
Rprof(NULL)
prof_hosvd_dense <- summaryRprof()
save(prof_hosvd_dense, file="prof_hosvd_dense.RData")
#################################################################
# hosvd (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".hosvd" 2.28 94.21 0.00 0.00
# "hosvd" 2.28 94.21 0.00 0.00
# "ttl" 1.88 77.69 0.00 0.00
# "as" 1.68 69.42 0.00 0.00
# "new" 1.38 57.02 0.06 2.48
# "initialize" 1.34 55.37 0.18 7.44
# "asMethod" 1.26 52.07 0.00 0.00
print("hosvd (Sparse)")
setSparse(TRUE)
Rprof()
hosvd(darr, ranks=c(2,3,4))
Rprof(NULL)
prof_hosvd_sparse <- summaryRprof()
save(prof_hosvd_sparse, file="prof_hosvd_sparse.RData")
#################################################################
# cp (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".cp" 5.52 97.87 0.00 0.00
# "cp" 5.52 97.87 0.00 0.00
# "as" 4.04 71.63 0.02 0.35
# "ttl" 3.24 57.45 0.00 0.00
# "new" 3.00 53.19 0.02 0.35
# "initialize" 2.96 52.48 0.42 7.45
print("cp (Dense)")
setSparse(FALSE)
Rprof()
cp(darr, num_components=2, max_iter=2)
Rprof(NULL)
prof_cp_dense <- summaryRprof()
save(prof_cp_dense, file="prof_cp_dense.RData")
#################################################################
# cp (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".cp" 4.76 97.94 0.00 0.00
# "cp" 4.76 97.94 0.00 0.00
# "as" 3.28 67.49 0.00 0.00
# "ttl" 2.86 58.85 0.00 0.00
setSparse(TRUE)
Rprof()
cp(darr, num_components=2, max_iter=2)
Rprof(NULL)
prof_cp_sparse <- summaryRprof()
save(prof_cp_sparse, file="prof_cp_sparse.RData")
#################################################################
# tucker (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".tucker" 12.82 99.07 0.00 0.00
# "tucker" 12.82 99.07 0.00 0.00
# "DelayedArray" 10.20 78.83 0.02 0.15
# "as" 9.36 72.33 0.02 0.15
# "<Anonymous>" 8.54 66.00 0.04 0.31
# ".super_BLOCK_mult" 8.44 65.22 0.00 0.00
# "do.call" 8.40 64.91 0.00 0.00
# ".svd" 8.34 64.45 0.00 0.00
# "realize" 8.32 64.30 0.00 0.00
# "asMethod" 8.26 63.83 0.00 0.00
# "withCallingHandlers" 8.02 61.98 0.02 0.15
# "suppressWarnings" 8.02 61.98 0.00 0.00
# "runIrlbaSVD" 8.00 61.82 0.00 0.00
# "standardGeneric" 7.88 60.90 0.12 0.93
# "mult" 7.68 59.35 0.00 0.00
# ".local" 7.42 57.34 0.00 0.00
# "drop" 7.38 57.03 0.00 0.00
# "writeHDF5Array" 6.78 52.40 0.00 0.00
print("tucker (Dense)")
setSparse(FALSE)
Rprof()
tucker(darr, ranks=c(2,2,2), max_iter=2)
Rprof(NULL)
prof_tucker_dense <- summaryRprof()
save(prof_tucker_dense, file="prof_tucker_dense.RData")
#################################################################
# tucker (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".tucker" 4.84 97.58 0.00 0.00
# "tucker" 4.84 97.58 0.00 0.00
# "ttl" 3.74 75.40 0.00 0.00
# "as" 3.58 72.18 0.04 0.81
# "asMethod" 2.60 52.42 0.00 0.00
# "new" 2.56 51.61 0.12 2.42
# "initialize" 2.50 50.40 0.32 6.45
print("tucker (Sparse)")
setSparse(TRUE)
Rprof()
tucker(darr, ranks=c(2,2,2), max_iter=2)
Rprof(NULL)
prof_tucker_sparse <- summaryRprof()
save(prof_tucker_sparse, file="prof_tucker_sparse.RData")
#################################################################
# mpca (Dense)
#################################################################
# total.time total.pct self.time self.pct
# ".mpca" 13.64 98.70 0.00 0.00
# "mpca" 13.64 98.70 0.00 0.00
# "DelayedArray" 11.40 82.49 0.00 0.00
# "<Anonymous>" 10.72 77.57 0.02 0.14
# "do.call" 10.54 76.27 0.00 0.00
# ".super_BLOCK_mult" 10.14 73.37 0.02 0.14
# ".svd" 10.04 72.65 0.00 0.00
# "as" 9.96 72.07 0.06 0.43
# "suppressWarnings" 9.82 71.06 0.02 0.14
# "runIrlbaSVD" 9.82 71.06 0.00 0.00
# "withCallingHandlers" 9.82 71.06 0.00 0.00
# "standardGeneric" 9.48 68.60 0.18 1.30
# "realize" 9.44 68.31 0.00 0.00
# "mult" 9.40 68.02 0.02 0.14
# "asMethod" 9.28 67.15 0.02 0.14
# "drop" 9.10 65.85 0.00 0.00
# ".local" 8.84 63.97 0.04 0.29
# "writeHDF5Array" 8.20 59.33 0.00 0.00
print("mpca (Dense)")
setSparse(FALSE)
Rprof()
mpca(darr, ranks=c(2,2), max_iter=2)
Rprof(NULL)
prof_mpca_dense <- summaryRprof()
save(prof_mpca_dense, file="prof_mpca_dense.RData")
#################################################################
# mpca (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".mpca" 3.70 95.36 0.00 0.00
# "mpca" 3.70 95.36 0.00 0.00
# "as" 2.60 67.01 0.00 0.00
# "ttl" 2.58 66.49 0.00 0.00
# "new" 2.12 54.64 0.00 0.00
# "initialize" 2.10 54.12 0.42 10.82
# "asMethod" 1.86 47.94 0.00 0.00
# ".reshapeIncNumbers1D" 1.74 44.85 0.00 0.00
# "DelayedArray" 1.74 44.85 0.00 0.00
# "realize" 1.74 44.85 0.00 0.00
print("mpca (Sparse)")
setSparse(TRUE)
Rprof()
mpca(darr, ranks=c(2,2), max_iter=2)
Rprof(NULL)
prof_mpca_sparse <- summaryRprof()
save(prof_mpca_sparse, file="prof_mpca_sparse.RData")
#################################################################
# pvd (Dense)
#################################################################
print("pvd (Dense)")
setSparse(FALSE)
Rprof()
pvd(darr, uranks=rep(2, dim(darr)[3]), wranks=rep(3, dim(darr)[3]), a=2, b=3)
Rprof(NULL)
prof_pvd_dense <- summaryRprof()
save(prof_pvd_dense, file="prof_pvd_dense.RData")
#################################################################
# pvd (Sparse)
#################################################################
# total.time total.pct self.time self.pct
# ".pvd" 15.64 98.24 0.00 0.00
# "pvd" 15.64 98.24 0.00 0.00
# "FUN" 13.30 83.54 0.08 0.50
# "lapply" 13.28 83.42 0.08 0.50
# "as" 13.10 82.29 0.12 0.75
# ".super_BLOCK_mult" 9.76 61.31 0.00 0.00
# "%*%" 9.76 61.31 0.00 0.00
# ".class1" 9.66 60.68 0.00 0.00
# "realize" 9.50 59.67 0.00 0.00
# "asMethod" 9.30 58.42 0.04 0.25
# "bplapply2" 9.16 57.54 0.00 0.00
print("pvd (Sparse)")
setSparse(TRUE)
Rprof()
pvd(darr, uranks=rep(2, dim(darr)[3]), wranks=rep(3, dim(darr)[3]), a=2, b=3)
Rprof(NULL)
prof_pvd_sparse <- summaryRprof()
save(prof_pvd_sparse, file="prof_pvd_sparse.RData")
# Sparse mode is effective in the most cases
# > prof_vec_dense$sampling.time
# [1] 0.12
# > prof_vec_sparse$sampling.time
# [1] 0.1
# > prof_unfold_dense$sampling.time
# [1] 0.64
# > prof_unfold_sparse$sampling.time
# [1] 0.2
# > prof_innerprod_dense$sampling.time
# [1] 0.04
# > prof_innerprod_sparse$sampling.time
# [1] 0.02
# > prof_hadamard_dense$sampling.time
# [1] 0.18
# > prof_hadamard_sparse$sampling.time
# [1] 0.3
# > prof_kronecker_dense$sampling.time
# [1] 13.18
# > prof_kronecker_sparse$sampling.time
# [1] 11
# > prof_khatri_rao_dense$sampling.time
# [1] 0.16
# > prof_khatri_rao_sparse$sampling.time
# [1] 0.16
# > prof_fold_dense$sampling.time
# [1] 0.24
# > prof_fold_sparse$sampling.time
# [1] 0.24
# > prof_modebind_list_dense$sampling.time
# [1] 0.4
# > prof_modebind_list_sparse$sampling.time
# [1] 0.38
# > prof_hosvd_dense$sampling.time
# [1] 11.06
# > prof_hosvd_sparse$sampling.time
# [1] 2.42
# > prof_cp_dense$sampling.time
# [1] 5.64
# > prof_cp_sparse$sampling.time
# [1] 4.86
# > prof_tucker_dense$sampling.time
# [1] 12.94
# > prof_tucker_sparse$sampling.time
# [1] 4.96
# > prof_mpca_dense$sampling.time
# [1] 13.82
# > prof_mpca_sparse$sampling.time
# [1] 3.88
# > prof_pvd_dense$sampling.time
# [1] 0
# > prof_pvd_sparse$sampling.time
# [1] 15.92
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.