matrixStats.benchmarks

---

allocMatrix() benchmarks

This report benchmark the performance of allocMatrix() against alternative methods.

Alternative methods

• matrix()
• matrix() special trick for NA

where

> allocMatrix_R <- function(nrow, ncol, value = NA) {
+     if (is.na(value) && !is.nan(value)) {
+         matrix(data = value[c()], nrow = nrow, ncol = ncol)
+     }     else {
+         matrix(data = value, nrow = nrow, ncol = ncol)
+     }
+ }

Data type “integer”

Data

> rmatrix <- function(nrow, ncol, mode = c("logical", "double", "integer", "index"), range = c(-100, 
+     +100), na_prob = 0) {
+     mode <- match.arg(mode)
+     n <- nrow * ncol
+     if (mode == "logical") {
+         x <- sample(c(FALSE, TRUE), size = n, replace = TRUE)
+     }     else if (mode == "index") {
+         x <- seq_len(n)
+         mode <- "integer"
+     }     else {
+         x <- runif(n, min = range[1], max = range[2])
+     }
+     storage.mode(x) <- mode
+     if (na_prob > 0) 
+         x[sample(n, size = na_prob * n)] <- NA
+     dim(x) <- c(nrow, ncol)
+     x
+ }
> rmatrices <- function(scale = 10, seed = 1, ...) {
+     set.seed(seed)
+     data <- list()
+     data[[1]] <- rmatrix(nrow = scale * 1, ncol = scale * 1, ...)
+     data[[2]] <- rmatrix(nrow = scale * 10, ncol = scale * 10, ...)
+     data[[3]] <- rmatrix(nrow = scale * 100, ncol = scale * 1, ...)
+     data[[4]] <- t(data[[3]])
+     data[[5]] <- rmatrix(nrow = scale * 10, ncol = scale * 100, ...)
+     data[[6]] <- t(data[[5]])
+     names(data) <- sapply(data, FUN = function(x) paste(dim(x), collapse = "x"))
+     data
+ }
> data <- rmatrices(mode = mode)
> values <- list(zero = 0, one = 1, `NA` = NA_real_)
> if (mode != "double") 
+     values <- lapply(values, FUN = function(x) {
+         storage.mode(x) <- mode
+         x
+     })

Results

10x10 matrix

> dim <- dim(data[["10x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x10+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.001448	0.0016640	0.0025898	0.0017960	0.0019500	0.077360
2	matrix	0.002029	0.0023095	0.0024605	0.0023985	0.0025870	0.004218
3	allocMatrix_R	0.003200	0.0034510	0.0038613	0.0035695	0.0036845	0.029245

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	matrix	1.401243	1.387921	0.9500471	1.335468	1.326667	0.0545243
3	allocMatrix_R	2.209945	2.073918	1.4909492	1.987472	1.889487	0.3780377

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x10+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["10x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x10+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.001274	0.0014780	0.0018879	0.0016845	0.0018500	0.009461
2	matrix	0.001716	0.0019715	0.0021898	0.0020535	0.0022090	0.003557
3	allocMatrix_R	0.002850	0.0030580	0.0037334	0.0031730	0.0033425	0.035492

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.346939	1.333897	1.159878	1.219056	1.194054	0.3759645
3	allocMatrix_R	2.237049	2.069012	1.977489	1.883645	1.806757	3.7514005

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x10+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["10x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x10+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.001298	0.001530	0.0019437	0.0016660	0.0017975	0.024871
2	matrix	0.001742	0.001938	0.0021533	0.0020895	0.0022535	0.004293
3	allocMatrix_R	0.003435	0.003639	0.0040764	0.0037755	0.0039200	0.026303

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.342065	1.266667	1.107868	1.254202	1.253686	0.1726107
3	allocMatrix_R	2.646379	2.378431	2.097254	2.266207	2.180807	1.0575771

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x10+NA data. Outliers are displayed as crosses. Times are in milliseconds.

100x100 matrix

> dim <- dim(data[["100x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x100+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.003411	0.003887	0.0043562	0.0040845	0.004372	0.011978
2	matrix	0.007401	0.007811	0.0081727	0.0080165	0.008244	0.011254
3	allocMatrix_R	0.008616	0.009111	0.0099560	0.0093295	0.009587	0.028301

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	2.169745	2.009519	1.876098	1.962664	1.885636	0.9395559
3	allocMatrix_R	2.525946	2.343967	2.285480	2.284123	2.192818	2.3627484

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x100+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["100x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x100+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.006351	0.0065825	0.0069510	0.0067895	0.0069815	0.015715
2	matrix	0.006848	0.0070225	0.0076476	0.0072860	0.0075340	0.030767
3	allocMatrix_R	0.008052	0.0083305	0.0087980	0.0085105	0.0087440	0.016906

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.078255	1.066844	1.100216	1.073128	1.079138	1.957811
3	allocMatrix_R	1.267832	1.265553	1.265719	1.253480	1.252453	1.075787

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x100+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["100x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x100+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.006841	0.0074550	0.0079802	0.007700	0.0080590	0.015584
2	matrix	0.007474	0.0080180	0.0085109	0.008256	0.0085495	0.020083
3	allocMatrix_R	0.012765	0.0131685	0.0141097	0.013508	0.0138895	0.049166

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.092530	1.075520	1.066498	1.072208	1.060864	1.288693
3	allocMatrix_R	1.865955	1.766398	1.768081	1.754286	1.723477	3.154902

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x100+NA data. Outliers are displayed as crosses. Times are in milliseconds.

1000x10 matrix

> dim <- dim(data[["1000x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x10+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.003550	0.0039295	0.0044064	0.0041925	0.0043985	0.014954
2	matrix	0.007361	0.0078105	0.0083108	0.0080100	0.0083135	0.021450
3	allocMatrix_R	0.008458	0.0090340	0.0097641	0.0092800	0.0094715	0.041340

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	2.073521	1.987658	1.886078	1.910555	1.890076	1.434399
3	allocMatrix_R	2.382535	2.299020	2.215883	2.213476	2.153348	2.764478

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x10+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["1000x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x10+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.006907	0.007534	0.0080022	0.007772	0.0080485	0.020347
2	matrix	0.007450	0.007983	0.0084510	0.008233	0.0085365	0.015048
3	allocMatrix_R	0.008656	0.009398	0.0100540	0.009539	0.0098610	0.040452

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.078616	1.059597	1.056083	1.059316	1.060632	0.7395685
3	allocMatrix_R	1.253221	1.247412	1.256401	1.227355	1.225197	1.9881064

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x10+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["1000x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x10+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.006890	0.007271	0.0081656	0.0076480	0.0080525	0.038914
2	matrix	0.007431	0.007795	0.0085203	0.0080530	0.0084235	0.030412
3	allocMatrix_R	0.012760	0.013192	0.0139413	0.0134745	0.0137235	0.026540

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.078520	1.072067	1.043441	1.052955	1.046073	0.7815182
3	allocMatrix_R	1.851959	1.814331	1.707317	1.761833	1.704253	0.6820168

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x10+NA data. Outliers are displayed as crosses. Times are in milliseconds.

10x1000 matrix

> dim <- dim(data[["10x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x1000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.003302	0.0039805	0.0053976	0.0050515	0.0060185	0.019499
2	matrix	0.007109	0.0080415	0.0116265	0.0096015	0.0154585	0.031072
3	allocMatrix_R	0.008372	0.0091940	0.0133690	0.0110080	0.0172700	0.021906

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	2.152938	2.020224	2.154031	1.900723	2.568497	1.593518
3	allocMatrix_R	2.535433	2.309760	2.476857	2.179155	2.869486	1.123442

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x1000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["10x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x1000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.006815	0.0073705	0.0079227	0.0076085	0.0080265	0.021806
2	matrix	0.007297	0.0079135	0.0084572	0.0081910	0.0086340	0.014578
3	allocMatrix_R	0.008513	0.0091230	0.0102696	0.0094655	0.0100450	0.040763

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.070726	1.073672	1.067470	1.076559	1.075687	0.6685316
3	allocMatrix_R	1.249156	1.237772	1.296223	1.244069	1.251480	1.8693479

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x1000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["10x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x1000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.006431	0.0070585	0.0076706	0.0075010	0.007912	0.014560
2	matrix	0.007081	0.0076155	0.0081192	0.0080455	0.008306	0.015418
3	allocMatrix_R	0.012309	0.0128055	0.0139109	0.0131620	0.013540	0.043787

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.101073	1.078912	1.058476	1.072590	1.049798	1.058929
3	allocMatrix_R	1.914010	1.814196	1.813535	1.754699	1.711325	3.007349

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+10x1000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

100x1000 matrix

> dim <- dim(data[["100x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x1000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.018195	0.0222325	0.1385829	0.025421	0.1747350	5.191896
3	allocMatrix_R	0.047737	0.0605130	0.1852318	0.172733	0.1997235	5.016253
2	matrix	0.046857	0.0585660	0.1444706	0.175918	0.1967650	0.303404

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	allocMatrix_R	2.623633	2.721826	1.336614	6.794894	1.143008	0.9661698
2	matrix	2.575268	2.634252	1.042485	6.920184	1.126077	0.0584380

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x1000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["100x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x1000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	matrix	0.046820	0.0525615	0.1613850	0.0589165	0.1805510	4.812541
1	allocMatrix	0.046546	0.0516115	0.1224808	0.1712130	0.1834835	0.228610
3	allocMatrix_R	0.047823	0.0530625	0.1829958	0.1765270	0.1889490	5.592455

	expr	min	lq	mean	median	uq	max
2	matrix	1.0000000	1.0000000	1.000000	1.000000	1.000000	1.000000
1	allocMatrix	0.9941478	0.9819259	0.758935	2.906028	1.016242	0.047503
3	allocMatrix_R	1.0214225	1.0095317	1.133908	2.996224	1.046513	1.162059

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x1000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["100x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x1000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.046322	0.0520500	0.1189115	0.1661990	0.1785575	0.213776
2	matrix	0.046489	0.0519965	0.1706100	0.1684770	0.1786830	5.419817
3	allocMatrix_R	0.075182	0.0797355	0.2062054	0.2022095	0.2136685	5.434947

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.0000000	1.000000	1.000000	1.000000	1.00000
2	matrix	1.003605	0.9989721	1.434764	1.013707	1.000703	25.35279
3	allocMatrix_R	1.623030	1.5319020	1.734109	1.216671	1.196637	25.42356

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+100x1000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

1000x100 matrix

> dim <- dim(data[["1000x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x100+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	matrix	0.047961	0.0518165	0.1209312	0.0759905	0.1853220	0.266259
3	allocMatrix_R	0.048024	0.0522910	0.2121506	0.1407865	0.1841975	9.042372
1	allocMatrix	0.018524	0.0225095	0.1661506	0.1659215	0.1806420	5.593055

	expr	min	lq	mean	median	uq	max
2	matrix	1.0000000	1.000000	1.000000	1.000000	1.0000000	1.00000
3	allocMatrix_R	1.0013136	1.009157	1.754308	1.852685	0.9939322	33.96081
1	allocMatrix	0.3862305	0.434408	1.373926	2.183451	0.9747467	21.00607

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x100+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["1000x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x100+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
3	allocMatrix_R	0.046637	0.0500700	0.1084039	0.063562	0.181088	0.224056
1	allocMatrix	0.045115	0.0522435	0.1226139	0.166599	0.181704	0.222297
2	matrix	0.045653	0.0529400	0.2307557	0.173530	0.183156	5.445768

	expr	min	lq	mean	median	uq	max
3	allocMatrix_R	1.0000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	allocMatrix	0.9673650	1.043409	1.131085	2.621047	1.003402	0.9921493
2	matrix	0.9789009	1.057320	2.128667	2.730090	1.011420	24.3053879

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x100+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["1000x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 int NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x100+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
3	allocMatrix_R	0.077507	0.0871955	0.2106748	0.1108355	0.235091	5.016970
1	allocMatrix	0.046450	0.0524495	0.1876490	0.1368740	0.195029	5.982023
2	matrix	0.047139	0.0533020	0.1418738	0.1731355	0.198120	0.364736

	expr	min	lq	mean	median	uq	max
3	allocMatrix_R	1.0000000	1.0000000	1.0000000	1.000000	1.0000000	1.0000000
1	allocMatrix	0.5993007	0.6015161	0.8907043	1.234929	0.8295894	1.1923577
2	matrix	0.6081902	0.6112930	0.6734254	1.562094	0.8427375	0.0727005

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on integer+1000x100+NA data. Outliers are displayed as crosses. Times are in milliseconds.

Data type “double”

Data

> rmatrix <- function(nrow, ncol, mode = c("logical", "double", "integer", "index"), range = c(-100, 
+     +100), na_prob = 0) {
+     mode <- match.arg(mode)
+     n <- nrow * ncol
+     if (mode == "logical") {
+         x <- sample(c(FALSE, TRUE), size = n, replace = TRUE)
+     }     else if (mode == "index") {
+         x <- seq_len(n)
+         mode <- "integer"
+     }     else {
+         x <- runif(n, min = range[1], max = range[2])
+     }
+     storage.mode(x) <- mode
+     if (na_prob > 0) 
+         x[sample(n, size = na_prob * n)] <- NA
+     dim(x) <- c(nrow, ncol)
+     x
+ }
> rmatrices <- function(scale = 10, seed = 1, ...) {
+     set.seed(seed)
+     data <- list()
+     data[[1]] <- rmatrix(nrow = scale * 1, ncol = scale * 1, ...)
+     data[[2]] <- rmatrix(nrow = scale * 10, ncol = scale * 10, ...)
+     data[[3]] <- rmatrix(nrow = scale * 100, ncol = scale * 1, ...)
+     data[[4]] <- t(data[[3]])
+     data[[5]] <- rmatrix(nrow = scale * 10, ncol = scale * 100, ...)
+     data[[6]] <- t(data[[5]])
+     names(data) <- sapply(data, FUN = function(x) paste(dim(x), collapse = "x"))
+     data
+ }
> data <- rmatrices(mode = mode)
> values <- list(zero = 0, one = 1, `NA` = NA_real_)
> if (mode != "double") 
+     values <- lapply(values, FUN = function(x) {
+         storage.mode(x) <- mode
+         x
+     })

Results

10x10 matrix

> dim <- dim(data[["10x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x10+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.001585	0.0018840	0.0030424	0.0020560	0.0029215	0.036657
2	matrix	0.002284	0.0024715	0.0031829	0.0026560	0.0036360	0.005946
3	allocMatrix_R	0.003437	0.0036975	0.0053181	0.0039705	0.0048815	0.086448

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.441010	1.311837	1.046157	1.291829	1.244566	0.1622064
3	allocMatrix_R	2.168454	1.962580	1.747990	1.931177	1.670888	2.3582945

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x10+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["10x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x10+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.001513	0.0017765	0.0019611	0.0018610	0.0019345	0.010277
2	matrix	0.002007	0.0022915	0.0023763	0.0023585	0.0024380	0.003266
3	allocMatrix_R	0.003267	0.0034805	0.0039440	0.0035560	0.0036835	0.035712

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.326504	1.289896	1.211704	1.267329	1.260274	0.317797
3	allocMatrix_R	2.159286	1.959189	2.011101	1.910801	1.904110	3.474944

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x10+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["10x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x10+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.001535	0.0020415	0.0034746	0.0029960	0.0035355	0.046115
2	matrix	0.001912	0.0029865	0.0040360	0.0037705	0.0042845	0.024676
3	allocMatrix_R	0.003554	0.0061695	0.0077345	0.0073445	0.0083615	0.035924

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.245603	1.462895	1.161570	1.258511	1.211851	0.535097
3	allocMatrix_R	2.315309	3.022043	2.226006	2.451435	2.365012	0.779009

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x10+NA data. Outliers are displayed as crosses. Times are in milliseconds.

100x100 matrix

> dim <- dim(data[["100x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x100+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.004949	0.0055870	0.0061887	0.0058835	0.0062295	0.017901
2	matrix	0.008467	0.0093770	0.0110810	0.0112415	0.0121420	0.026045
3	allocMatrix_R	0.009511	0.0107405	0.0130282	0.0125240	0.0136460	0.043152

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.710851	1.678360	1.790509	1.910682	1.949113	1.454947
3	allocMatrix_R	1.921802	1.922409	2.105152	2.128665	2.190545	2.410592

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x100+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["100x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x100+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.007808	0.0085005	0.0091886	0.0088755	0.0094375	0.021502
2	matrix	0.008458	0.0090275	0.0096667	0.0092375	0.0096645	0.033331
3	allocMatrix_R	0.009341	0.0102525	0.0108400	0.0105795	0.0109545	0.024375

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.083248	1.061996	1.052028	1.040786	1.024053	1.550135
3	allocMatrix_R	1.196337	1.206106	1.179722	1.191989	1.160742	1.133616

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x100+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["100x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x100+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.008388	0.009414	0.0117857	0.012381	0.0133490	0.020894
2	matrix	0.008846	0.009864	0.0122597	0.013193	0.0139255	0.019238
3	allocMatrix_R	0.012482	0.013430	0.0155517	0.015339	0.0163010	0.045200

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.054602	1.047801	1.040215	1.065584	1.043187	0.9207428
3	allocMatrix_R	1.488078	1.426599	1.319541	1.238914	1.221140	2.1633005

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x100+NA data. Outliers are displayed as crosses. Times are in milliseconds.

1000x10 matrix

> dim <- dim(data[["1000x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x10+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.005057	0.005736	0.0076880	0.0063770	0.0082540	0.044611
3	allocMatrix_R	0.009650	0.010955	0.0145061	0.0119185	0.0166140	0.040787
2	matrix	0.008378	0.009840	0.0148882	0.0130940	0.0182375	0.058953

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	allocMatrix_R	1.908246	1.909868	1.886842	1.868982	2.012842	0.9142812
2	matrix	1.656713	1.715481	1.936552	2.053317	2.209535	1.3214902

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x10+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["1000x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x10+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.008235	0.0097530	0.0116499	0.0121455	0.0128385	0.027389
2	matrix	0.008418	0.0095580	0.0122272	0.0127845	0.0135555	0.041526
3	allocMatrix_R	0.009796	0.0108485	0.0131687	0.0139365	0.0148415	0.020145

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.0000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.022222	0.9800062	1.049555	1.052612	1.055848	1.5161561
3	allocMatrix_R	1.189557	1.1123244	1.130375	1.147462	1.156015	0.7355143

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x10+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["1000x10"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x10+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.008582	0.0095490	0.0116630	0.0123205	0.0130560	0.019537
2	matrix	0.009041	0.0100005	0.0125355	0.0128610	0.0140485	0.039084
3	allocMatrix_R	0.013356	0.0145565	0.0160547	0.0154915	0.0163015	0.047047

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.053484	1.047282	1.074806	1.043870	1.076019	2.000512
3	allocMatrix_R	1.556281	1.524401	1.376547	1.257376	1.248583	2.408097

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x10+NA data. Outliers are displayed as crosses. Times are in milliseconds.

10x1000 matrix

> dim <- dim(data[["10x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x1000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.005412	0.0059245	0.0065421	0.0063155	0.0067060	0.020415
2	matrix	0.009188	0.0100430	0.0120108	0.0120025	0.0133005	0.022162
3	allocMatrix_R	0.010548	0.0116715	0.0136441	0.0135505	0.0145345	0.043294

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.697709	1.695164	1.835919	1.900483	1.983373	1.085574
3	allocMatrix_R	1.949002	1.970040	2.085576	2.145594	2.167387	2.120696

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x1000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["10x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x1000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.008301	0.0092825	0.0109210	0.0098470	0.0124750	0.026631
2	matrix	0.008974	0.0098150	0.0117544	0.0104330	0.0132365	0.043284
3	allocMatrix_R	0.010225	0.0111835	0.0130230	0.0118615	0.0147635	0.023901

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.081075	1.057366	1.076303	1.059511	1.061042	1.6253239
3	allocMatrix_R	1.231779	1.204794	1.192467	1.204580	1.183447	0.8974879

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x1000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["10x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x1000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.008524	0.0094095	0.0113736	0.0100385	0.0128535	0.043693
2	matrix	0.008907	0.0097660	0.0113957	0.0102705	0.0131300	0.017067
3	allocMatrix_R	0.013156	0.0139745	0.0152347	0.0149395	0.0159595	0.028209

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.044932	1.037887	1.001948	1.023111	1.021512	0.3906118
3	allocMatrix_R	1.543407	1.485148	1.339478	1.488220	1.241646	0.6456183

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+10x1000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

100x1000 matrix

> dim <- dim(data[["100x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x1000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
3	allocMatrix_R	0.064921	0.0718620	0.2540681	0.2107430	0.3213915	5.234302
2	matrix	0.065602	0.0714770	0.3208140	0.3075490	0.3225895	5.430003
1	allocMatrix	0.035362	0.0422275	0.2766638	0.3257345	0.3469525	5.538367

	expr	min	lq	mean	median	uq	max
3	allocMatrix_R	1.0000000	1.0000000	1.000000	1.000000	1.000000	1.000000
2	matrix	1.0104897	0.9946425	1.262708	1.459356	1.003728	1.037388
1	allocMatrix	0.5446928	0.5876193	1.088936	1.545648	1.079532	1.058091

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x1000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["100x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x1000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	matrix	0.068200	0.0719105	0.2622136	0.3064355	0.3179015	5.731292
1	allocMatrix	0.066682	0.0719535	0.2720804	0.3076740	0.3178365	5.309507
3	allocMatrix_R	0.069095	0.0737530	0.3093317	0.3096100	0.3218795	4.914369

	expr	min	lq	mean	median	uq	max
2	matrix	1.0000000	1.000000	1.000000	1.000000	1.0000000	1.0000000
1	allocMatrix	0.9777419	1.000598	1.037629	1.004042	0.9997955	0.9264066
3	allocMatrix_R	1.0131232	1.025622	1.179694	1.010359	1.0125133	0.8574627

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x1000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["100x1000"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x1000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.065926	0.0715965	0.2511258	0.3016935	0.3143690	5.387141
2	matrix	0.066325	0.0718785	0.2707839	0.3066210	0.3239570	5.234148
3	allocMatrix_R	0.079111	0.0885815	0.3493736	0.3387680	0.3515995	5.617906

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	matrix	1.006052	1.003939	1.078280	1.016333	1.030499	0.9716003
3	allocMatrix_R	1.199997	1.237232	1.391229	1.122888	1.118429	1.0428363

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+100x1000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

1000x100 matrix

> dim <- dim(data[["1000x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 0

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x100+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	matrix	0.063275	0.0708785	0.2103157	0.3072135	0.319865	0.389493
3	allocMatrix_R	0.063570	0.0720110	0.2703309	0.3097320	0.325722	5.256226
1	allocMatrix	0.033570	0.0405345	0.3580392	0.3254795	0.347529	5.459805

	expr	min	lq	mean	median	uq	max
2	matrix	1.0000000	1.0000000	1.000000	1.000000	1.000000	1.00000
3	allocMatrix_R	1.0046622	1.0159780	1.285358	1.008198	1.018311	13.49505
1	allocMatrix	0.5305413	0.5718871	1.702390	1.059457	1.086486	14.01772

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x100+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["1000x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num 1

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x100+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	matrix	0.067016	0.0761700	0.3090076	0.3039075	0.3108040	5.260457
3	allocMatrix_R	0.069564	0.0744480	0.2651395	0.3039485	0.3166690	5.421525
1	allocMatrix	0.066568	0.0728625	0.2701209	0.3055060	0.3171215	5.258883

	expr	min	lq	mean	median	uq	max
2	matrix	1.000000	1.0000000	1.0000000	1.000000	1.000000	1.0000000
3	allocMatrix_R	1.038021	0.9773927	0.8580358	1.000135	1.018870	1.0306186
1	allocMatrix	0.993315	0.9565774	0.8741562	1.005260	1.020326	0.9997008

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x100+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> dim <- dim(data[["1000x100"]])
> nrow <- dim[1L]
> ncol <- dim[2L]
> str(value)
 num NA

> stats <- microbenchmark(allocMatrix = allocMatrix(nrow = nrow, ncol = ncol, value = value), matrix = matrix(data = value, 
+     nrow = nrow, ncol = ncol), allocMatrix_R = allocMatrix_R(nrow = nrow, ncol = ncol, value = value), 
+     unit = "ms")

Table: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x100+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocMatrix	0.067364	0.0716220	0.2112774	0.2988155	0.310120	0.373895
2	matrix	0.066557	0.0708965	0.3020821	0.3005430	0.311621	5.112225
3	allocMatrix_R	0.076113	0.0845335	0.3286843	0.3325165	0.350639	5.301227

	expr	min	lq	mean	median	uq	max
1	allocMatrix	1.0000000	1.0000000	1.00000	1.000000	1.000000	1.00000
2	matrix	0.9880203	0.9898704	1.42979	1.005781	1.004840	13.67289
3	allocMatrix_R	1.1298765	1.1802728	1.55570	1.112782	1.130656	14.17838

Figure: Benchmarking of allocMatrix(), matrix() and allocMatrix_R() on double+1000x100+NA data. Outliers are displayed as crosses. Times are in milliseconds.

Appendix

Session information

R version 4.1.1 Patched (2021-08-10 r80727)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.5 LTS

Matrix products: default
BLAS:   /home/hb/software/R-devel/R-4-1-branch/lib/R/lib/libRblas.so
LAPACK: /home/hb/software/R-devel/R-4-1-branch/lib/R/lib/libRlapack.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] microbenchmark_1.4-7   matrixStats_0.60.0     ggplot2_3.3.5         
[4] knitr_1.33             R.devices_2.17.0       R.utils_2.10.1        
[7] R.oo_1.24.0            R.methodsS3_1.8.1-9001 history_0.0.1-9000    

loaded via a namespace (and not attached):
 [1] Biobase_2.52.0          httr_1.4.2              splines_4.1.1          
 [4] bit64_4.0.5             network_1.17.1          assertthat_0.2.1       
 [7] highr_0.9               stats4_4.1.1            blob_1.2.2             
[10] GenomeInfoDbData_1.2.6  robustbase_0.93-8       pillar_1.6.2           
[13] RSQLite_2.2.8           lattice_0.20-44         glue_1.4.2             
[16] digest_0.6.27           XVector_0.32.0          colorspace_2.0-2       
[19] Matrix_1.3-4            XML_3.99-0.7            pkgconfig_2.0.3        
[22] zlibbioc_1.38.0         genefilter_1.74.0       purrr_0.3.4            
[25] ergm_4.1.2              xtable_1.8-4            scales_1.1.1           
[28] tibble_3.1.4            annotate_1.70.0         KEGGREST_1.32.0        
[31] farver_2.1.0            generics_0.1.0          IRanges_2.26.0         
[34] ellipsis_0.3.2          cachem_1.0.6            withr_2.4.2            
[37] BiocGenerics_0.38.0     survival_3.2-13         magrittr_2.0.1         
[40] crayon_1.4.1            statnet.common_4.5.0    memoise_2.0.0          
[43] laeken_0.5.1            fansi_0.5.0             R.cache_0.15.0         
[46] MASS_7.3-54             R.rsp_0.44.0            progressr_0.8.0        
[49] tools_4.1.1             lifecycle_1.0.0         S4Vectors_0.30.0       
[52] trust_0.1-8             munsell_0.5.0           tabby_0.0.1-9001       
[55] AnnotationDbi_1.54.1    Biostrings_2.60.2       compiler_4.1.1         
[58] GenomeInfoDb_1.28.1     rlang_0.4.11            grid_4.1.1             
[61] RCurl_1.98-1.4          cwhmisc_6.6             rstudioapi_0.13        
[64] rappdirs_0.3.3          startup_0.15.0          labeling_0.4.2         
[67] bitops_1.0-7            base64enc_0.1-3         boot_1.3-28            
[70] gtable_0.3.0            DBI_1.1.1               R6_2.5.1               
[73] lpSolveAPI_5.5.2.0-17.7 rle_0.9.2               dplyr_1.0.7            
[76] fastmap_1.1.0           bit_4.0.4               utf8_1.2.2             
[79] parallel_4.1.1          Rcpp_1.0.7              vctrs_0.3.8            
[82] png_0.1-7               DEoptimR_1.0-9          tidyselect_1.1.1       
[85] xfun_0.25               coda_0.19-4            

Total processing time was 35.78 secs.

Reproducibility

To reproduce this report, do:

html <- matrixStats:::benchmark('allocMatrix')

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 22:04:57 (+0200 UTC). Powered by RSP.