matrixStats.benchmarks

---

colSums2() and rowSums2() benchmarks

This report benchmark the performance of colSums2() and rowSums2() against alternative methods.

Alternative methods

• apply() + sum()
• colSums() and rowSums()
• .colSums() and .rowSums()

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"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5301799 283.2    8529671 455.6  8529671 455.6
Vcells 10537069  80.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5293402 282.7    8529671 455.6  8529671 455.6
Vcells 10509835  80.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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	0.002024	0.0023590	0.0029285	0.0029400	0.0032180	0.013506
2	.colSums	0.002832	0.0032685	0.0038752	0.0036150	0.0041135	0.020128
3	colSums	0.005209	0.0058285	0.0065548	0.0063255	0.0069795	0.017134
4	apply+sum	0.033529	0.0351655	0.0374598	0.0362565	0.0373940	0.105952

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.colSums	1.399209	1.385545	1.323262	1.229592	1.278278	1.490301
3	colSums	2.573617	2.470750	2.238267	2.151531	2.168894	1.268621
4	apply+sum	16.565711	14.906952	12.791366	12.332143	11.620261	7.844810

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.001982	0.0022625	0.0028599	0.0028210	0.0031610	0.013250
2	.rowSums	0.004119	0.0044760	0.0048506	0.0047345	0.0049525	0.017034
3	rowSums	0.006220	0.0067815	0.0075147	0.0073010	0.0076885	0.030091
4	apply+sum	0.033936	0.0355055	0.0376751	0.0365405	0.0376265	0.095222

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	2.078204	1.978342	1.696050	1.678306	1.566751	1.285585
3	rowSums	3.138244	2.997348	2.627587	2.588089	2.432300	2.271019
4	apply+sum	17.122099	15.693039	13.173392	12.953031	11.903353	7.186566

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+10x10 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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	1.982	2.2625	2.85994	2.821	3.161	13.250
1	colSums2	2.024	2.3590	2.92852	2.940	3.218	13.506

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.000000	1.000000	1.00000	1.000000	1.000000	1.000000
1	colSums2	1.021191	1.042652	1.02398	1.042184	1.018032	1.019321

Figure: Benchmarking of colSums2() and rowSums2() on integer+10x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x100 integer matrix

> X <- data[["100x100"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5292003 282.7    8529671 455.6  8529671 455.6
Vcells 10126681  77.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5291979 282.7    8529671 455.6  8529671 455.6
Vcells 10131694  77.3   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.013267	0.0143255	0.0155111	0.0148430	0.0160555	0.030816
3	colSums	0.015436	0.0169870	0.0190825	0.0179955	0.0193405	0.064016
1	colSums2	0.025059	0.0265480	0.0281877	0.0274060	0.0284910	0.044906
4	apply+sum	0.191798	0.2036240	0.2223822	0.2116170	0.2340260	0.360976

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.163488	1.185788	1.230243	1.212390	1.204603	2.077362
1	colSums2	1.888822	1.853199	1.817256	1.846392	1.774532	1.457230
4	apply+sum	14.456772	14.214094	14.336941	14.257024	14.576064	11.713915

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.023951	0.0261330	0.0290982	0.0284265	0.0316665	0.047280
2	.rowSums	0.039585	0.0416140	0.0459872	0.0441170	0.0500685	0.065475
3	rowSums	0.041789	0.0447305	0.0487430	0.0465710	0.0515080	0.069365
4	apply+sum	0.185463	0.1977110	0.2203234	0.2092150	0.2359210	0.362197

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.652749	1.592393	1.580413	1.551967	1.581119	1.384835
3	rowSums	1.744771	1.711648	1.675117	1.638295	1.626577	1.467111
4	apply+sum	7.743435	7.565568	7.571709	7.359858	7.450176	7.660681

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+100x100 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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	25.059	26.548	28.18769	27.4060	28.4910	44.906
2	rowSums2	23.951	26.133	29.09824	28.4265	31.6665	47.280

	expr	min	lq	mean	median	uq	max
1	colSums2	1.0000000	1.0000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2	0.9557843	0.9843679	1.032303	1.037236	1.111456	1.052866

Figure: Benchmarking of colSums2() and rowSums2() on integer+100x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x10 integer matrix

> X <- data[["1000x10"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5292748 282.7    8529671 455.6  8529671 455.6
Vcells 10130440  77.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5292742 282.7    8529671 455.6  8529671 455.6
Vcells 10135483  77.4   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.012877	0.0135665	0.0145759	0.0143975	0.0152645	0.021049
3	colSums	0.015109	0.0163470	0.0178206	0.0170805	0.0185490	0.041538
1	colSums2	0.025325	0.0264235	0.0283863	0.0283695	0.0298660	0.041995
4	apply+sum	0.102768	0.1065335	0.1145748	0.1130425	0.1193445	0.200489

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.173332	1.204953	1.222605	1.186352	1.215172	1.973395
1	colSums2	1.966685	1.947702	1.947480	1.970446	1.956566	1.995107
4	apply+sum	7.980741	7.852689	7.860550	7.851537	7.818435	9.524871

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.020458	0.0225050	0.0250921	0.0245700	0.0265265	0.047041
4	apply+sum	0.084019	0.0930460	0.1043740	0.1011610	0.1138445	0.205897
2	.rowSums	0.138914	0.1472935	0.1689857	0.1663425	0.1850230	0.219647
3	rowSums	0.140435	0.1539695	0.1707097	0.1683390	0.1811930	0.229037

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	apply+sum	4.106902	4.134459	4.159634	4.117257	4.291727	4.376969
2	.rowSums	6.790204	6.544923	6.734613	6.770147	6.975025	4.669267
3	rowSums	6.864552	6.841569	6.803318	6.851404	6.830641	4.868880

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+1000x10 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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
2	rowSums2	20.458	22.5050	25.09212	24.5700	26.5265	47.041
1	colSums2	25.325	26.4235	28.38634	28.3695	29.8660	41.995

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.000000	1.000000	1.000000	1.00000	1.000000	1.0000000
1	colSums2	1.237902	1.174117	1.131285	1.15464	1.125893	0.8927319

Figure: Benchmarking of colSums2() and rowSums2() on integer+1000x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

10x1000 integer matrix

> X <- data[["10x1000"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5292989 282.7    8529671 455.6  8529671 455.6
Vcells 10131283  77.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5292965 282.7    8529671 455.6  8529671 455.6
Vcells 10136296  77.4   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.012056	0.012832	0.0149051	0.0139830	0.0165545	0.024744
3	colSums	0.013839	0.015257	0.0184374	0.0169405	0.0196615	0.074987
1	colSums2	0.022568	0.023974	0.0274259	0.0251180	0.0291710	0.050615
4	apply+sum	0.920811	0.967966	1.0634032	1.0228625	1.1259850	1.525506

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.147893	1.188981	1.236990	1.211507	1.187683	3.030512
1	colSums2	1.871931	1.868298	1.840035	1.796324	1.762119	2.045546
4	apply+sum	76.377820	75.433759	71.345066	73.150433	68.016853	61.651552

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.021958	0.0239985	0.0264151	0.0247900	0.0266345	0.059772
2	.rowSums	0.024765	0.0266265	0.0285517	0.0272280	0.0284600	0.047468
3	rowSums	0.026245	0.0286380	0.0312012	0.0297175	0.0317650	0.055039
4	apply+sum	0.925215	0.9716065	1.0564796	1.0201945	1.1103690	1.496538

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.rowSums	1.127835	1.109507	1.080886	1.098346	1.068539	0.7941511
3	rowSums	1.195236	1.193325	1.181189	1.198770	1.192626	0.9208158
4	apply+sum	42.135668	40.486135	39.995306	41.153469	41.689125	25.0374423

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+10x1000 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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
2	rowSums2	21.958	23.9985	26.41509	24.790	26.6345	59.772
1	colSums2	22.568	23.9740	27.42585	25.118	29.1710	50.615

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.00000	1.0000000	1.000000	1.000000	1.000000	1.0000000
1	colSums2	1.02778	0.9989791	1.038264	1.013231	1.095234	0.8468012

Figure: Benchmarking of colSums2() and rowSums2() on integer+10x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x1000 integer matrix

> X <- data[["100x1000"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5293207 282.7    8529671 455.6  8529671 455.6
Vcells 10131856  77.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5293195 282.7    8529671 455.6  8529671 455.6
Vcells 10181889  77.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.087684	0.0943470	0.1008148	0.0977915	0.1020245	0.178827
3	colSums	0.089896	0.0950085	0.1019444	0.1002585	0.1049210	0.146086
1	colSums2	0.191433	0.2071520	0.2167812	0.2134215	0.2193430	0.297765
4	apply+sum	1.428837	1.5786385	1.6531785	1.6204870	1.6569505	2.598360

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	colSums	1.025227	1.007011	1.011205	1.025227	1.028390	0.8169124
1	colSums2	2.183215	2.195639	2.150291	2.182414	2.149905	1.6651009
4	apply+sum	16.295299	16.732260	16.398171	16.570837	16.240712	14.5300206

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.191744	0.2088630	0.2179392	0.2133365	0.2204645	0.309175
2	.rowSums	0.220700	0.2400595	0.2532652	0.2491540	0.2536810	0.379838
3	rowSums	0.222796	0.2470615	0.2562721	0.2548550	0.2576245	0.330766
4	apply+sum	1.439896	1.5791325	1.6386681	1.6127200	1.6362765	2.669264

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.151014	1.149363	1.162091	1.167892	1.150666	1.228553
3	rowSums	1.161945	1.182888	1.175888	1.194615	1.168553	1.069834
4	apply+sum	7.509471	7.560614	7.518922	7.559513	7.421950	8.633505

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+100x1000 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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
2	rowSums2	191.744	208.863	217.9392	213.3365	220.4645	309.175
1	colSums2	191.433	207.152	216.7811	213.4215	219.3430	297.765

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
1	colSums2	0.998378	0.991808	0.9946861	1.000398	0.994913	0.9630953

Figure: Benchmarking of colSums2() and rowSums2() on integer+100x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x100 integer matrix

> X <- data[["1000x100"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5293431 282.7    8529671 455.6  8529671 455.6
Vcells 10132554  77.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5293419 282.7    8529671 455.6  8529671 455.6
Vcells 10182587  77.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.080745	0.0854250	0.0962083	0.0902415	0.0983220	0.188483
3	colSums	0.083270	0.0872815	0.1002490	0.0929945	0.1059230	0.225327
1	colSums2	0.183159	0.1911685	0.2122587	0.2039955	0.2314095	0.314580
4	apply+sum	0.631666	0.6720135	0.7621986	0.7066705	0.8105770	1.694147

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.031271	1.021732	1.041999	1.030507	1.077307	1.195477
1	colSums2	2.268363	2.237852	2.206241	2.260551	2.353588	1.669010
4	apply+sum	7.822974	7.866708	7.922378	7.830882	8.244106	8.988328

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.185045	0.1899330	0.2064439	0.2020740	0.2082815	0.300979
2	.rowSums	0.319892	0.3284885	0.3629335	0.3512650	0.3600065	0.813243
3	rowSums	0.322223	0.3401980	0.3646800	0.3600730	0.3694095	0.517022
4	apply+sum	0.643562	0.6826685	0.7342289	0.7079275	0.7289560	1.131492

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.728725	1.729497	1.758025	1.738299	1.728461	2.701992
3	rowSums	1.741322	1.791147	1.766485	1.781887	1.773607	1.717801
4	apply+sum	3.477867	3.594260	3.556554	3.503308	3.499860	3.759372

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+1000x100 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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
2	rowSums2	185.045	189.9330	206.4439	202.0740	208.2815	300.979
1	colSums2	183.159	191.1685	212.2587	203.9955	231.4095	314.580

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.0000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colSums2	0.9898079	1.006505	1.028167	1.009509	1.111042	1.045189

Figure: Benchmarking of colSums2() and rowSums2() 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"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5293674 282.8    8529671 455.6  8529671 455.6
Vcells 10248485  78.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5293641 282.8    8529671 455.6  8529671 455.6
Vcells 10248583  78.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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	0.002102	0.0024065	0.0030994	0.0029520	0.0032165	0.017842
2	.colSums	0.002778	0.0033325	0.0038123	0.0036850	0.0040910	0.013051
3	colSums	0.005430	0.0059150	0.0066628	0.0063655	0.0072170	0.015898
4	apply+sum	0.034034	0.0365380	0.0386263	0.0374980	0.0380990	0.110486

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.colSums	1.321598	1.384791	1.230041	1.248306	1.271879	0.7314763
3	colSums	2.583254	2.457926	2.149721	2.156335	2.243743	0.8910436
4	apply+sum	16.191246	15.183046	12.462676	12.702575	11.844862	6.1924672

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.001940	0.0022655	0.0028561	0.0028985	0.0031365	0.013258
2	.rowSums	0.002398	0.0026455	0.0031198	0.0029570	0.0031550	0.017511
3	rowSums	0.004646	0.0050370	0.0058332	0.0056080	0.0060780	0.028458
4	apply+sum	0.034348	0.0353945	0.0371595	0.0359440	0.0372065	0.091068

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.236082	1.167733	1.092327	1.020183	1.005898	1.320787
3	rowSums	2.394845	2.223350	2.042344	1.934794	1.937829	2.146478
4	apply+sum	17.705155	15.623262	13.010378	12.400897	11.862426	6.868909

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+10x10 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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
2	rowSums2	1.940	2.2655	2.85614	2.8985	3.1365	13.258
1	colSums2	2.102	2.4065	3.09936	2.9520	3.2165	17.842

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colSums2	1.083505	1.062238	1.085157	1.018458	1.025506	1.345754

Figure: Benchmarking of colSums2() and rowSums2() on double+10x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x100 double matrix

> X <- data[["100x100"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5293893 282.8    8529671 455.6  8529671 455.6
Vcells 10249501  78.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5293869 282.8    8529671 455.6  8529671 455.6
Vcells 10259514  78.3   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.010384	0.0115435	0.0124507	0.0119520	0.0132230	0.018895
3	colSums	0.012993	0.0142790	0.0159926	0.0150795	0.0162580	0.035340
1	colSums2	0.020164	0.0209950	0.0226975	0.0219650	0.0235610	0.037274
4	apply+sum	0.206727	0.2129235	0.2348531	0.2243070	0.2453505	0.375133

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.251252	1.236973	1.284470	1.261672	1.229524	1.870336
1	colSums2	1.941834	1.818772	1.822985	1.837768	1.781820	1.972691
4	apply+sum	19.908224	18.445315	18.862626	18.767319	18.554829	19.853559

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.022182	0.0242220	0.0265808	0.0259425	0.0287055	0.040805
2	.rowSums	0.027795	0.0291870	0.0316074	0.0304490	0.0335525	0.048271
3	rowSums	0.029620	0.0324980	0.0350392	0.0336310	0.0364825	0.054420
4	apply+sum	0.197381	0.2095255	0.2303125	0.2213215	0.2425360	0.372191

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.253043	1.204979	1.189107	1.173711	1.168853	1.182968
3	rowSums	1.335317	1.341673	1.318215	1.296367	1.270924	1.333660
4	apply+sum	8.898251	8.650215	8.664615	8.531232	8.449113	9.121211

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+100x100 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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
1	colSums2	20.164	20.995	22.69746	21.9650	23.5610	37.274
2	rowSums2	22.182	24.222	26.58081	25.9425	28.7055	40.805

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2	1.100079	1.153703	1.171092	1.181083	1.218348	1.094731

Figure: Benchmarking of colSums2() and rowSums2() on double+100x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x10 double matrix

> X <- data[["1000x10"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5294116 282.8    8529671 455.6  8529671 455.6
Vcells 10250597  78.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5294092 282.8    8529671 455.6  8529671 455.6
Vcells 10260610  78.3   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.012371	0.0134835	0.0144271	0.0141420	0.0151645	0.025370
3	colSums	0.015010	0.0162415	0.0175628	0.0171425	0.0182715	0.035068
1	colSums2	0.018749	0.0201340	0.0219062	0.0217815	0.0228090	0.034523
4	apply+sum	0.111090	0.1168800	0.1254827	0.1231010	0.1331115	0.233988

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.213321	1.204546	1.217350	1.212169	1.204886	1.382262
1	colSums2	1.515561	1.493232	1.518406	1.540199	1.504105	1.360780
4	apply+sum	8.979872	8.668373	8.697717	8.704639	8.777836	9.223019

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.022056	0.0231260	0.0248843	0.0241820	0.0261350	0.041828
2	.rowSums	0.030313	0.0311025	0.0345391	0.0338760	0.0367065	0.051703
3	rowSums	0.032743	0.0340905	0.0369900	0.0362275	0.0394180	0.052757
4	apply+sum	0.110896	0.1160200	0.1261189	0.1234515	0.1337685	0.223937

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.374365	1.344915	1.387987	1.400877	1.404496	1.236086
3	rowSums	1.484539	1.474120	1.486478	1.498118	1.508246	1.261284
4	apply+sum	5.027929	5.016864	5.068206	5.105099	5.118366	5.353758

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+1000x10 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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	18.749	20.134	21.90617	21.7815	22.809	34.523
2	rowSums2	22.056	23.126	24.88433	24.1820	26.135	41.828

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000
2	rowSums2	1.176383	1.148604	1.135951	1.110208	1.14582	1.211598

Figure: Benchmarking of colSums2() and rowSums2() on double+1000x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

10x1000 double matrix

> X <- data[["10x1000"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5294339 282.8    8529671 455.6  8529671 455.6
Vcells 10250745  78.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5294315 282.8    8529671 455.6  8529671 455.6
Vcells 10260758  78.3   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.008916	0.0097405	0.0114750	0.0108600	0.0124010	0.024709
3	colSums	0.010672	0.0119185	0.0155951	0.0133570	0.0158335	0.097707
1	colSums2	0.021120	0.0222750	0.0255460	0.0232775	0.0279855	0.045889
4	apply+sum	0.927717	0.9840360	1.0874877	1.0289875	1.1331390	1.933843

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.196949	1.223602	1.359048	1.229926	1.276792	3.954308
1	colSums2	2.368775	2.286844	2.226230	2.143416	2.256713	1.857177
4	apply+sum	104.050808	101.025204	94.770169	94.750230	91.374809	78.264721

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.022778	0.024432	0.0266944	0.025203	0.0277680	0.056972
2	.rowSums	0.023898	0.025582	0.0272317	0.026342	0.0271710	0.045543
3	rowSums	0.025444	0.027818	0.0300013	0.028727	0.0306545	0.053482
4	apply+sum	0.935441	0.981507	1.0554225	1.028925	1.0806680	1.509979

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.0000000	1.0000000
2	.rowSums	1.049170	1.047069	1.020128	1.045193	0.9785004	0.7993927
3	rowSums	1.117043	1.138589	1.123880	1.139825	1.1039506	0.9387418
4	apply+sum	41.067741	40.173011	39.537255	40.825517	38.9177470	26.5038791

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+10x1000 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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	21.120	22.275	25.54599	23.2775	27.9855	45.889
2	rowSums2	22.778	24.432	26.69438	25.2030	27.7680	56.972

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.000000	1.000000	1.0000000	1.000000
2	rowSums2	1.078504	1.096835	1.044954	1.082719	0.9922281	1.241518

Figure: Benchmarking of colSums2() and rowSums2() on double+10x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x1000 double matrix

> X <- data[["100x1000"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5294557 282.8    8529671 455.6  8529671 455.6
Vcells 10252015  78.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5294545 282.8    8529671 455.6  8529671 455.6
Vcells 10352048  79.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.063484	0.0693695	0.0772934	0.072477	0.0811175	0.166440
3	colSums	0.065582	0.0714795	0.0793142	0.074465	0.0833640	0.156227
1	colSums2	0.146856	0.1590045	0.1747343	0.163916	0.1797135	0.491681
4	apply+sum	1.465818	1.6159060	1.9389410	1.671202	1.8663865	19.263417

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	colSums	1.033048	1.030417	1.026144	1.027429	1.027694	0.9386385
1	colSums2	2.313276	2.292139	2.260663	2.261628	2.215471	2.9541036
4	apply+sum	23.089566	23.294186	25.085470	23.058377	23.008432	115.7379056

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.175242	0.1908420	0.2015122	0.1959230	0.2029835	0.286708
2	.rowSums	0.209709	0.2293750	0.2414981	0.2340870	0.2414530	0.369516
3	rowSums	0.211997	0.2359765	0.2497697	0.2390785	0.2600210	0.342801
4	apply+sum	1.496664	1.6280940	1.9198785	1.6703115	1.8006130	19.846470

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.196682	1.201911	1.198429	1.194791	1.189520	1.288823
3	rowSums	1.209739	1.236502	1.239477	1.220268	1.280996	1.195645
4	apply+sum	8.540555	8.531109	9.527354	8.525347	8.870736	69.221891

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+100x1000 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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	146.856	159.0045	174.7343	163.916	179.7135	491.681
2	rowSums2	175.242	190.8420	201.5122	195.923	202.9835	286.708

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.00000	1.000000	1.000000	1.000000	1.0000000
2	rowSums2	1.193291	1.20023	1.153249	1.195265	1.129484	0.5831179

Figure: Benchmarking of colSums2() and rowSums2() on double+100x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x100 double matrix

> X <- data[["1000x100"]]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5294793 282.8    8529671 455.6  8529671 455.6
Vcells 10252184  78.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5294769 282.8    8529671 455.6  8529671 455.6
Vcells 10352197  79.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.077817	0.080332	0.0900894	0.0831115	0.095637	0.154776
3	colSums	0.080178	0.083281	0.0934385	0.0883645	0.100575	0.138330
1	colSums2	0.132999	0.137269	0.1531161	0.1462835	0.162858	0.231035
4	apply+sum	0.687649	0.720354	0.9015292	0.7735955	0.906300	8.450792

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	colSums	1.030340	1.036710	1.037175	1.063204	1.051633	0.8937432
1	colSums2	1.709125	1.708771	1.699601	1.760087	1.702876	1.4927056
4	apply+sum	8.836745	8.967211	10.007045	9.307924	9.476458	54.6001447

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	0.168968	0.1741515	0.1976026	0.1851140	0.2080825	0.385057
2	.rowSums	0.214875	0.2209335	0.2498737	0.2349080	0.2704140	0.379430
3	rowSums	0.216937	0.2298130	0.2508185	0.2432355	0.2649335	0.354560
4	apply+sum	0.709047	0.7667290	0.9330494	0.8030805	0.9409255	8.399390

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.rowSums	1.271691	1.268628	1.264527	1.268991	1.299552	0.9853866
3	rowSums	1.283894	1.319615	1.269308	1.313977	1.273214	0.9207987
4	apply+sum	4.196339	4.402655	4.721848	4.338302	4.521887	21.8133679

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+1000x100 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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	132.999	137.2690	153.1161	146.2835	162.8580	231.035
2	rowSums2	168.968	174.1515	197.6026	185.1140	208.0825	385.057

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2	1.270446	1.268688	1.290541	1.265447	1.277693	1.666661

Figure: Benchmarking of colSums2() and rowSums2() 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_0.60.1     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] 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               markdown_1.1           
[73] R6_2.5.1                lpSolveAPI_5.5.2.0-17.7 rle_0.9.2              
[76] dplyr_1.0.7             fastmap_1.1.0           bit_4.0.4              
[79] utf8_1.2.2              parallel_4.1.1          Rcpp_1.0.7             
[82] vctrs_0.3.8             png_0.1-7               DEoptimR_1.0-9         
[85] tidyselect_1.1.1        xfun_0.25               coda_0.19-4            

Total processing time was 25.15 secs.

Reproducibility

To reproduce this report, do:

html <- matrixStats:::benchmark('colSums2')

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 19:09:32 (+0200 UTC). Powered by RSP.