matrixStats.benchmarks

---

colSums2() and rowSums2() benchmarks on subsetted computation

This report benchmark the performance of colSums2() and rowSums2() on subsetted computation.

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)

Results

10x10 integer matrix

> X <- data[["10x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5314824 283.9    7916910 422.9  7916910 422.9
Vcells 10551177  80.5   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302496 283.2    7916910 422.9  7916910 422.9
Vcells 10510317  80.2   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+10x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.002316	0.002403	0.0042154	0.0024930	0.0025880	0.171107
2	colSums2(X, rows, cols)	0.002642	0.002754	0.0028602	0.0027970	0.0028915	0.005203
3	colSums2(X[rows, cols])	0.003259	0.003513	0.0037776	0.0036155	0.0037805	0.009327

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	colSums2(X, rows, cols)	1.140760	1.146067	0.6785098	1.121941	1.117272	0.0304079
3	colSums2(X[rows, cols])	1.407167	1.461923	0.8961491	1.450261	1.460781	0.0545098

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on integer+10x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.001869	0.0019430	0.0020549	0.0019900	0.002142	0.003934
2	rowSums2(X, cols, rows)	0.002119	0.0022120	0.0037345	0.0022715	0.002373	0.143996
3	rowSums2(X[cols, rows])	0.002585	0.0027965	0.0029551	0.0028880	0.003017	0.004617

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.133761	1.138446	1.817317	1.141457	1.107843	36.602949
3	rowSums2(X[cols, rows])	1.383093	1.439269	1.438059	1.451256	1.408497	1.173615

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+10x10 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+10x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	1.869	1.943	2.05493	1.990	2.142	3.934
1	colSums2_X_S	2.316	2.403	4.21537	2.493	2.588	171.107

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
1	colSums2_X_S	1.239165	1.236747	2.051345	1.252764	1.208217	43.49441

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+10x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x100 integer matrix

> X <- data[["100x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5301206 283.2    7916910 422.9  7916910 422.9
Vcells 10179412  77.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5301200 283.2    7916910 422.9  7916910 422.9
Vcells 10184495  77.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.008729	0.0092500	0.0094428	0.0093860	0.0095560	0.013562
2	colSums2(X, rows, cols)	0.010685	0.0112615	0.0116479	0.0114020	0.0116605	0.027005
3	colSums2(X[rows, cols])	0.019133	0.0197825	0.0207284	0.0199755	0.0203620	0.069989

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.224081	1.217459	1.233526	1.214788	1.220228	1.991226
3	colSums2(X[rows, cols])	2.191889	2.138649	2.195152	2.128223	2.130808	5.160670

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on integer+100x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.009558	0.0100625	0.0103888	0.0102950	0.0104620	0.021480
2	rowSums2(X, cols, rows)	0.011635	0.0119750	0.0124275	0.0121325	0.0122575	0.039039
3	rowSums2(X[cols, rows])	0.020905	0.0215570	0.0219882	0.0217545	0.0219440	0.028624

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.217305	1.190062	1.196237	1.178485	1.171621	1.817458
3	rowSums2(X[cols, rows])	2.187173	2.142311	2.116524	2.113113	2.097496	1.332588

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+100x100 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+100x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	8.729	9.2500	9.44280	9.386	9.556	13.562
2	rowSums2_X_S	9.558	10.0625	10.38881	10.295	10.462	21.480

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2_X_S	1.094971	1.087838	1.100183	1.096846	1.094809	1.583837

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+100x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x10 integer matrix

> X <- data[["1000x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5301949 283.2    7916910 422.9  7916910 422.9
Vcells 10183453  77.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5301943 283.2    7916910 422.9  7916910 422.9
Vcells 10188536  77.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+1000x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.010044	0.0103990	0.0122987	0.0106630	0.0144510	0.021642
2	colSums2(X, rows, cols)	0.012258	0.0126535	0.0159935	0.0129625	0.0206970	0.038640
3	colSums2(X[rows, cols])	0.020871	0.0213755	0.0280079	0.0227695	0.0339855	0.070511

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.220430	1.216800	1.300424	1.215652	1.432219	1.785417
3	colSums2(X[rows, cols])	2.077957	2.055534	2.277310	2.135375	2.351775	3.258063

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on integer+1000x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.011398	0.0119945	0.0129188	0.0123815	0.0129920	0.020887
2	rowSums2(X, cols, rows)	0.013514	0.0140985	0.0155152	0.0144615	0.0151190	0.033319
3	rowSums2(X[cols, rows])	0.025292	0.0260055	0.0290323	0.0270035	0.0281085	0.117424

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.185647	1.175414	1.200978	1.167993	1.163716	1.595203
3	rowSums2(X[cols, rows])	2.218986	2.168119	2.247286	2.180956	2.163524	5.621870

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+1000x10 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+1000x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	10.044	10.3990	12.29868	10.6630	14.451	21.642
2	rowSums2_X_S	11.398	11.9945	12.91881	12.3815	12.992	20.887

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.0000000	1.0000000
2	rowSums2_X_S	1.134807	1.153428	1.050423	1.161165	0.8990381	0.9651141

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+1000x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

10x1000 integer matrix

> X <- data[["10x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302154 283.2    7916910 422.9  7916910 422.9
Vcells 10184297  77.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302148 283.2    7916910 422.9  7916910 422.9
Vcells 10189380  77.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+10x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.011157	0.0115725	0.0121712	0.0117770	0.0120475	0.033054
2	colSums2(X, rows, cols)	0.013578	0.0140480	0.0145518	0.0143615	0.0147585	0.027031
3	colSums2(X[rows, cols])	0.024245	0.0250570	0.0258912	0.0257265	0.0259075	0.040396

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.216994	1.213912	1.195585	1.219453	1.225026	0.817783
3	colSums2(X[rows, cols])	2.173075	2.165219	2.127241	2.184470	2.150446	1.222121

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on integer+10x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.011518	0.011811	0.0120629	0.0119535	0.012196	0.016302
2	rowSums2(X, cols, rows)	0.014313	0.014660	0.0151715	0.0148175	0.015115	0.040562
3	rowSums2(X[cols, rows])	0.023530	0.023992	0.0245609	0.0242060	0.024589	0.032016

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.242664	1.241216	1.257698	1.239595	1.239341	2.488161
3	rowSums2(X[cols, rows])	2.042889	2.031327	2.036068	2.025014	2.016153	1.963931

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+10x1000 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+10x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	11.157	11.5725	12.17125	11.7770	12.0475	33.054
2	rowSums2_X_S	11.518	11.8110	12.06290	11.9535	12.1960	16.302

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.000000
2	rowSums2_X_S	1.032356	1.020609	0.9910979	1.014987	1.012326	0.493193

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+10x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x1000 integer matrix

> X <- data[["100x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302364 283.2    7916910 422.9  7916910 422.9
Vcells 10206969  77.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302358 283.2    7916910 422.9  7916910 422.9
Vcells 10257052  78.3   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.058661	0.060829	0.0704060	0.0684605	0.0747285	0.114031
2	colSums2(X, rows, cols)	0.076577	0.079850	0.0942961	0.0924130	0.0992465	0.159023
3	colSums2(X[rows, cols])	0.143722	0.155078	0.1763339	0.1734750	0.1892045	0.262634

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.305416	1.312696	1.339320	1.349873	1.328094	1.394559
3	colSums2(X[rows, cols])	2.450044	2.549409	2.504531	2.533943	2.531892	2.303181

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on integer+100x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.059031	0.0685180	0.0797687	0.0751715	0.0811330	0.164440
2	rowSums2(X, cols, rows)	0.093292	0.1005085	0.1218268	0.1083040	0.1213630	0.248732
3	rowSums2(X[cols, rows])	0.138484	0.1569265	0.1879376	0.1732365	0.1911615	0.356114

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.580390	1.466892	1.527250	1.440759	1.495853	1.512600
3	rowSums2(X[cols, rows])	2.345954	2.290296	2.356031	2.304550	2.356150	2.165617

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+100x1000 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+100x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	58.661	60.829	70.40595	68.4605	74.7285	114.031
2	rowSums2_X_S	59.031	68.518	79.76871	75.1715	81.1330	164.440

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2_X_S	1.006307	1.126403	1.132983	1.098027	1.085704	1.442064

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+100x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x100 integer matrix

> X <- data[["1000x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302577 283.2    7916910 422.9  7916910 422.9
Vcells 10207748  77.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302571 283.2    7916910 422.9  7916910 422.9
Vcells 10257831  78.3   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.064606	0.0683840	0.0735184	0.069482	0.0776080	0.121728
2	colSums2(X, rows, cols)	0.070671	0.0749850	0.0816940	0.080333	0.0867945	0.107117
3	colSums2(X[rows, cols])	0.141236	0.1507725	0.1644976	0.163143	0.1747045	0.247660

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colSums2(X, rows, cols)	1.093877	1.096528	1.111204	1.156170	1.118370	0.8799701
3	colSums2(X[rows, cols])	2.186113	2.204792	2.237502	2.347989	2.251115	2.0345360

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on integer+1000x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.065778	0.071136	0.0773413	0.074036	0.082757	0.104619
2	rowSums2(X, cols, rows)	0.087792	0.094717	0.1037324	0.098624	0.112170	0.192951
3	rowSums2(X[cols, rows])	0.142763	0.152797	0.1675265	0.161947	0.181091	0.232181

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.334671	1.331492	1.341230	1.332109	1.355414	1.844321
3	rowSums2(X[cols, rows])	2.170376	2.147956	2.166069	2.187409	2.188226	2.219301

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+1000x100 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+1000x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	64.606	68.384	73.51841	69.482	77.608	121.728
2	rowSums2_X_S	65.778	71.136	77.34126	74.036	82.757	104.619

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowSums2_X_S	1.018141	1.040243	1.051999	1.065542	1.066346	0.8594489

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+1000x100 data (original and transposed). 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)

Results

10x10 double matrix

> X <- data[["10x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302794 283.2    7916910 422.9  7916910 422.9
Vcells 10298849  78.6   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302779 283.2    7916910 422.9  7916910 422.9
Vcells 10299017  78.6   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+10x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.002341	0.002414	0.0027210	0.0024780	0.002578	0.022386
2	colSums2(X, rows, cols)	0.002714	0.002840	0.0029690	0.0029170	0.003032	0.005705
3	colSums2(X[rows, cols])	0.003303	0.003574	0.0037463	0.0036495	0.003712	0.010135

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colSums2(X, rows, cols)	1.159334	1.176471	1.091147	1.177159	1.176106	0.2548468
3	colSums2(X[rows, cols])	1.410935	1.480530	1.376795	1.472760	1.439876	0.4527383

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on double+10x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.002333	0.0024105	0.0025451	0.0024800	0.0025760	0.005639
2	rowSums2(X, cols, rows)	0.002684	0.0028410	0.0033580	0.0029210	0.0030395	0.041357
3	rowSums2(X[cols, rows])	0.003303	0.0035485	0.0037562	0.0036225	0.0037710	0.010750

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.150450	1.178594	1.319384	1.177823	1.179930	7.334102
3	rowSums2(X[cols, rows])	1.415774	1.472101	1.475881	1.460686	1.463898	1.906366

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+10x10 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+10x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	2.341	2.4140	2.72100	2.478	2.578	22.386
2	rowSums2_X_S	2.333	2.4105	2.54509	2.480	2.576	5.639

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.0000000	1.0000000	1.000000	1.000000	1.0000000	1.0000000
2	rowSums2_X_S	0.9965827	0.9985501	0.935351	1.000807	0.9992242	0.2518985

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+10x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x100 double matrix

> X <- data[["100x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302991 283.3    7916910 422.9  7916910 422.9
Vcells 10304792  78.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5302985 283.3    7916910 422.9  7916910 422.9
Vcells 10314875  78.7   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.010459	0.0107905	0.0112199	0.0109635	0.0111370	0.025726
2	colSums2(X, rows, cols)	0.016901	0.0173970	0.0176763	0.0175880	0.0177435	0.022675
3	colSums2(X[rows, cols])	0.026682	0.0272295	0.0282483	0.0274205	0.0277075	0.076606

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.615929	1.612251	1.575439	1.604232	1.593203	0.881404
3	colSums2(X[rows, cols])	2.551104	2.523470	2.517692	2.501072	2.487878	2.977766

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on double+100x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.009046	0.0093890	0.0097173	0.0096400	0.0098320	0.019483
2	rowSums2(X, cols, rows)	0.014577	0.0149175	0.0155950	0.0151315	0.0155135	0.043478
3	rowSums2(X[cols, rows])	0.022414	0.0234785	0.0243372	0.0239535	0.0243425	0.038641

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.61143	1.588827	1.604869	1.569658	1.577858	2.231587
3	rowSums2(X[cols, rows])	2.47778	2.500639	2.504516	2.484803	2.475844	1.983319

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+100x100 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+100x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	9.046	9.3890	9.71732	9.6400	9.832	19.483
1	colSums2_X_S	10.459	10.7905	11.21993	10.9635	11.137	25.726

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	1.000000	1.00000	1.000000	1.000000	1.00000	1.000000
1	colSums2_X_S	1.156202	1.14927	1.154632	1.137293	1.13273	1.320433

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+100x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x10 double matrix

> X <- data[["1000x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5303191 283.3    7916910 422.9  7916910 422.9
Vcells 10306189  78.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5303185 283.3    7916910 422.9  7916910 422.9
Vcells 10316272  78.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+1000x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.010561	0.011050	0.0119243	0.0113745	0.0118005	0.025584
2	colSums2(X, rows, cols)	0.016280	0.016798	0.0180738	0.0173060	0.0179170	0.029592
3	colSums2(X[rows, cols])	0.026334	0.026877	0.0291187	0.0277250	0.0286385	0.072998

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.541521	1.520181	1.515706	1.521473	1.518325	1.156660
3	colSums2(X[rows, cols])	2.493514	2.432308	2.441956	2.437470	2.426889	2.853268

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on double+1000x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.011542	0.0118265	0.0120980	0.0119990	0.0122380	0.017745
2	rowSums2(X, cols, rows)	0.017504	0.0179935	0.0183176	0.0181665	0.0184215	0.025606
3	rowSums2(X[cols, rows])	0.031572	0.0318735	0.0328445	0.0321000	0.0324665	0.084338

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.516548	1.521456	1.514102	1.514001	1.505270	1.442998
3	rowSums2(X[cols, rows])	2.735401	2.695092	2.714867	2.675223	2.652925	4.752775

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+1000x10 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+1000x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	10.561	11.0500	11.92432	11.3745	11.8005	25.584
2	rowSums2_X_S	11.542	11.8265	12.09800	11.9990	12.2380	17.745

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowSums2_X_S	1.092889	1.070272	1.014565	1.054903	1.037075	0.6935976

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+1000x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

10x1000 double matrix

> X <- data[["10x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5303396 283.3    7916910 422.9  7916910 422.9
Vcells 10306325  78.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5303390 283.3    7916910 422.9  7916910 422.9
Vcells 10316408  78.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+10x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.011208	0.012267	0.0127463	0.0125435	0.0128485	0.034210
2	colSums2(X, rows, cols)	0.020316	0.021219	0.0219282	0.0216850	0.0219920	0.040049
3	colSums2(X[rows, cols])	0.030857	0.032037	0.0331032	0.0326290	0.0330240	0.047968

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.812634	1.729763	1.720353	1.728784	1.711639	1.170681
3	colSums2(X[rows, cols])	2.753123	2.611641	2.597081	2.601268	2.570261	1.402163

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on double+10x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.012283	0.0128835	0.0139251	0.0133335	0.0136495	0.046460
2	rowSums2(X, cols, rows)	0.020440	0.0209255	0.0217699	0.0213985	0.0218075	0.057897
3	rowSums2(X[cols, rows])	0.029424	0.0303700	0.0308753	0.0307965	0.0312235	0.038420

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowSums2(X, cols, rows)	1.664089	1.624209	1.563357	1.604867	1.597678	1.2461687
3	rowSums2(X[cols, rows])	2.395506	2.357279	2.217241	2.309709	2.287520	0.8269479

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+10x1000 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+10x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	11.208	12.2670	12.74633	12.5435	12.8485	34.21
2	rowSums2_X_S	12.283	12.8835	13.92511	13.3335	13.6495	46.46

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.00000	1.000000	1.000000	1.000000
2	rowSums2_X_S	1.095914	1.050257	1.09248	1.062981	1.062342	1.358082

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+10x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x1000 double matrix

> X <- data[["100x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5303606 283.3    7916910 422.9  7916910 422.9
Vcells 10351779  79.0   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5303600 283.3    7916910 422.9  7916910 422.9
Vcells 10451862  79.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.060942	0.0631560	0.0699083	0.065866	0.0764510	0.108410
2	colSums2(X, rows, cols)	0.109678	0.1134325	0.1258434	0.121479	0.1416550	0.233528
3	colSums2(X[rows, cols])	0.178424	0.1864040	0.2074802	0.201984	0.2316995	0.250504

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.799711	1.796069	1.800121	1.844335	1.852886	2.154119
3	colSums2(X[rows, cols])	2.927767	2.951485	2.967890	3.066590	3.030693	2.310709

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on double+100x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.069882	0.0748485	0.0816588	0.0805325	0.0840435	0.190951
2	rowSums2(X, cols, rows)	0.126393	0.1359265	0.1443958	0.1411055	0.1533300	0.267493
3	rowSums2(X[cols, rows])	0.196918	0.2227935	0.2346010	0.2323730	0.2481025	0.282891

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.808663	1.816022	1.768282	1.752156	1.824412	1.400846
3	rowSums2(X[cols, rows])	2.817864	2.976593	2.872942	2.885456	2.952072	1.481485

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+100x1000 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+100x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	60.942	63.1560	69.90832	65.8660	76.4510	108.410
2	rowSums2_X_S	69.882	74.8485	81.65882	80.5325	84.0435	190.951

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2_X_S	1.146697	1.185137	1.168084	1.222672	1.099312	1.761378

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+100x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x100 double matrix

> X <- data[["1000x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5303819 283.3    7916910 422.9  7916910 422.9
Vcells 10351923  79.0   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5303813 283.3    7916910 422.9  7916910 422.9
Vcells 10452006  79.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	0.066988	0.0691450	0.0748006	0.071450	0.0796690	0.111544
2	colSums2(X, rows, cols)	0.101823	0.1055185	0.1159303	0.113096	0.1265495	0.169818
3	colSums2(X[rows, cols])	0.182219	0.1896335	0.2110076	0.200158	0.2371660	0.351424

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.520018	1.526047	1.549857	1.582869	1.588441	1.522431
3	colSums2(X[rows, cols])	2.720174	2.742548	2.820933	2.801372	2.976892	3.150542

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on double+1000x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	0.067781	0.0691705	0.0760977	0.0723850	0.0815540	0.130793
2	rowSums2(X, cols, rows)	0.111252	0.1142600	0.1260092	0.1199190	0.1344080	0.283172
3	rowSums2(X[cols, rows])	0.187477	0.1974990	0.2160575	0.2081705	0.2356565	0.267566

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.641345	1.651860	1.655887	1.656683	1.648086	2.165039
3	rowSums2(X[cols, rows])	2.765923	2.855249	2.839212	2.875879	2.889576	2.045721

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+1000x100 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+1000x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	66.988	69.1450	74.80063	71.450	79.669	111.544
2	rowSums2_X_S	67.781	69.1705	76.09769	72.385	81.554	130.793

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.00000	1.000000	1.00000	1.000000
2	rowSums2_X_S	1.011838	1.000369	1.01734	1.013086	1.02366	1.172569

Figure: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+1000x100 data (original and transposed). 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     mime_0.11               survival_3.2-13        
[40] magrittr_2.0.1          crayon_1.4.1            statnet.common_4.5.0   
[43] memoise_2.0.0           laeken_0.5.1            fansi_0.5.0            
[46] R.cache_0.15.0          MASS_7.3-54             R.rsp_0.44.0           
[49] progressr_0.8.0         tools_4.1.1             lifecycle_1.0.0        
[52] S4Vectors_0.30.0        trust_0.1-8             munsell_0.5.0          
[55] tabby_0.0.1-9001        AnnotationDbi_1.54.1    Biostrings_2.60.2      
[58] compiler_4.1.1          GenomeInfoDb_1.28.1     rlang_0.4.11           
[61] grid_4.1.1              RCurl_1.98-1.4          cwhmisc_6.6            
[64] rstudioapi_0.13         rappdirs_0.3.3          startup_0.15.0         
[67] labeling_0.4.2          bitops_1.0-7            base64enc_0.1-3        
[70] boot_1.3-28             gtable_0.3.0            DBI_1.1.1              
[73] markdown_1.1            R6_2.5.1                lpSolveAPI_5.5.2.0-17.7
[76] rle_0.9.2               dplyr_1.0.7             fastmap_1.1.0          
[79] bit_4.0.4               utf8_1.2.2              parallel_4.1.1         
[82] Rcpp_1.0.7              vctrs_0.3.8             png_0.1-7              
[85] DEoptimR_1.0-9          tidyselect_1.1.1        xfun_0.25              
[88] coda_0.19-4            

Total processing time was 25.02 secs.

Reproducibility

To reproduce this report, do:

html <- matrixStats:::benchmark('colRowSums2_subset')

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 22:29:34 (+0200 UTC). Powered by RSP.