matrixStats.benchmarks

---

colMeans2() and rowMeans2() benchmarks

This report benchmark the performance of colMeans2() and rowMeans2() against alternative methods.

Alternative methods

• apply() + mean()
• .colMeans() and .rowMeans()
• colMeans() and rowMeans()

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  5243331 280.1    8529671 455.6  8529671 455.6
Vcells 10144820  77.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5234941 279.6    8529671 455.6  8529671 455.6
Vcells 10117596  77.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	colMeans2	0.001992	0.0023665	0.0031423	0.0030835	0.0033555	0.016988
2	.colMeans	0.002875	0.0032820	0.0038932	0.0036305	0.0042620	0.018913
3	colMeans	0.005486	0.0060380	0.0069369	0.0064100	0.0075990	0.018538
4	apply+mean	0.069440	0.0719295	0.0740808	0.0730880	0.0742025	0.149651

	expr	min	lq	mean	median	uq	max
1	colMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.colMeans	1.443273	1.386858	1.238933	1.177396	1.270153	1.113315
3	colMeans	2.754016	2.551447	2.207549	2.078807	2.264640	1.091241
4	apply+mean	34.859438	30.394887	23.575056	23.702935	22.113694	8.809218

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.002010	0.0023115	0.0029323	0.0028950	0.0032590	0.013166
2	.rowMeans	0.004297	0.0045735	0.0049267	0.0048095	0.0050785	0.012926
3	rowMeans	0.006422	0.0070070	0.0079371	0.0077705	0.0081275	0.035648
4	apply+mean	0.069812	0.0727465	0.0743678	0.0735045	0.0742985	0.134797

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.rowMeans	2.137811	1.978585	1.680165	1.661313	1.558300	0.9817712
3	rowMeans	3.195025	3.031365	2.706806	2.684111	2.493863	2.7075801
4	apply+mean	34.732338	31.471555	25.361946	25.390155	22.797944	10.2382652

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on integer+10x10 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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	rowMeans2	2.010	2.3115	2.93226	2.8950	3.2590	13.166
1	colMeans2	1.992	2.3665	3.14234	3.0835	3.3555	16.988

	expr	min	lq	mean	median	uq	max
2	rowMeans2	1.0000000	1.000000	1.000000	1.000000	1.00000	1.000000
1	colMeans2	0.9910448	1.023794	1.071644	1.065112	1.02961	1.290293

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5233542 279.6    8529671 455.6  8529671 455.6
Vcells 9734433  74.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5233518 279.6    8529671 455.6  8529671 455.6
Vcells 9739446  74.4   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.011740	0.0129460	0.0147181	0.014228	0.015126	0.030573
1	colMeans2	0.013761	0.0151360	0.0170067	0.016602	0.017374	0.032148
3	colMeans	0.014046	0.0159465	0.0178452	0.017108	0.019011	0.036727
4	apply+mean	0.404897	0.4364410	0.4947802	0.481770	0.534527	0.727672

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colMeans2	1.172146	1.169164	1.155498	1.166854	1.148618	1.051516
3	colMeans	1.196422	1.231770	1.212467	1.202418	1.256843	1.201289
4	apply+mean	34.488671	33.712421	33.617103	33.860697	35.338292	23.801132

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.015711	0.0175810	0.0208816	0.0204045	0.0239075	0.035039
2	.rowMeans	0.034387	0.0360530	0.0425171	0.0403155	0.0461885	0.064760
3	rowMeans	0.036265	0.0400645	0.0463726	0.0447190	0.0495415	0.073816
4	apply+mean	0.393155	0.4262150	0.4861259	0.4767065	0.5237225	0.729521

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowMeans	2.188721	2.050680	2.036106	1.975814	1.931967	1.848226
3	rowMeans	2.308255	2.278852	2.220744	2.191624	2.072216	2.106681
4	apply+mean	25.024187	24.242933	23.280133	23.362812	21.906201	20.820257

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on integer+100x100 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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	colMeans2	13.761	15.136	17.00674	16.6020	17.3740	32.148
2	rowMeans2	15.711	17.581	20.88158	20.4045	23.9075	35.039

	expr	min	lq	mean	median	uq	max
1	colMeans2	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000
2	rowMeans2	1.141705	1.161535	1.227841	1.229039	1.37605	1.089928

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5234288 279.6    8529671 455.6  8529671 455.6
Vcells 9738176  74.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5234282 279.6    8529671 455.6  8529671 455.6
Vcells 9743219  74.4   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.013163	0.0143130	0.0155026	0.0154285	0.0163975	0.022736
3	colMeans	0.015870	0.0175615	0.0189664	0.0186625	0.0197770	0.035407
1	colMeans2	0.015938	0.0173365	0.0188227	0.0189070	0.0195330	0.031747
4	apply+mean	0.141983	0.1481960	0.1607782	0.1592475	0.1725340	0.251937

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colMeans	1.205652	1.226962	1.223431	1.209612	1.206098	1.557310
1	colMeans2	1.210818	1.211242	1.214164	1.225459	1.191218	1.396332
4	apply+mean	10.786523	10.353944	10.371052	10.321645	10.521970	11.080973

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.015900	0.0175550	0.0198808	0.0189100	0.0208750	0.038201
4	apply+mean	0.118695	0.1301820	0.1446521	0.1382570	0.1604360	0.259639
2	.rowMeans	0.146888	0.1554555	0.1738174	0.1708555	0.1906365	0.219340
3	rowMeans	0.148658	0.1588325	0.1761765	0.1715380	0.1868535	0.229172

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	apply+mean	7.465094	7.415665	7.275974	7.311317	7.685557	6.796654
2	.rowMeans	9.238239	8.855340	8.742980	9.035193	9.132287	5.741734
3	rowMeans	9.349560	9.047707	8.861643	9.071285	8.951066	5.999110

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on integer+1000x10 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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	colMeans2	15.938	17.3365	18.82268	18.907	19.533	31.747
2	rowMeans2	15.900	17.5550	19.88079	18.910	20.875	38.201

	expr	min	lq	mean	median	uq	max
1	colMeans2	1.0000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMeans2	0.9976158	1.012603	1.056215	1.000159	1.068704	1.203295

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5234528 279.6    8529671 455.6  8529671 455.6
Vcells 9739019  74.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5234504 279.6    8529671 455.6  8529671 455.6
Vcells 9744032  74.4   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.013135	0.014760	0.0158475	0.0157035	0.0163910	0.022401
3	colMeans	0.015766	0.017293	0.0202043	0.0187220	0.0211025	0.054562
1	colMeans2	0.016770	0.019353	0.0213923	0.0205510	0.0222930	0.038533
4	apply+mean	3.089859	3.492587	3.6795149	3.5659015	3.8066240	8.120902

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colMeans	1.200305	1.171612	1.274921	1.192218	1.287444	2.435695
1	colMeans2	1.276741	1.311179	1.349890	1.308689	1.360076	1.720146
4	apply+mean	235.238599	236.625135	232.183111	227.076862	232.238668	362.524084

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.017299	0.019108	0.0219312	0.0201595	0.0217275	0.052354
2	.rowMeans	0.026678	0.029969	0.0313585	0.0308465	0.0318540	0.049340
3	rowMeans	0.029382	0.032307	0.0356464	0.0336205	0.0363340	0.077620
4	apply+mean	3.112655	3.484507	3.6993173	3.6053875	3.8273165	8.703466

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.00000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.rowMeans	1.54217	1.568401	1.429861	1.530122	1.466068	0.9424304
3	rowMeans	1.69848	1.690758	1.625374	1.667725	1.672259	1.4825992
4	apply+mean	179.93266	182.358567	168.678445	178.843101	176.150800	166.2426176

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on integer+10x1000 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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	rowMeans2	17.299	19.108	21.93118	20.1595	21.7275	52.354
1	colMeans2	16.770	19.353	21.39234	20.5510	22.2930	38.533

	expr	min	lq	mean	median	uq	max
2	rowMeans2	1.0000000	1.000000	1.0000000	1.00000	1.000000	1.0000000
1	colMeans2	0.9694202	1.012822	0.9754304	1.01942	1.026027	0.7360087

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5234746 279.6    8529671 455.6  8529671 455.6
Vcells 9739590  74.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5234734 279.6    8529671 455.6  8529671 455.6
Vcells 9789623  74.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.088949	0.1015470	0.1064333	0.1039070	0.1072950	0.192288
3	colMeans	0.090834	0.1030705	0.1075519	0.1054680	0.1104115	0.146975
1	colMeans2	0.119515	0.1331360	0.1387987	0.1375605	0.1435235	0.185253
4	apply+mean	3.759292	4.2359465	4.5301877	4.3567710	4.5313390	18.450519

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	colMeans	1.021192	1.015003	1.010510	1.015023	1.029046	0.7643483
1	colMeans2	1.343635	1.311078	1.304091	1.323881	1.337653	0.9634143
4	apply+mean	42.263454	41.714147	42.563631	41.929524	42.232527	95.9525243

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.144974	0.1653780	0.1728834	0.1683720	0.1758220	0.289812
2	.rowMeans	0.222891	0.2537550	0.2598834	0.2559765	0.2628260	0.383208
3	rowMeans	0.229851	0.2565955	0.2605143	0.2596985	0.2671795	0.310940
4	apply+mean	3.751237	4.1897070	4.5080480	4.3244395	4.5011175	18.660145

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowMeans	1.537455	1.534394	1.503229	1.520303	1.494841	1.322264
3	rowMeans	1.585464	1.551570	1.506878	1.542409	1.519602	1.072902
4	apply+mean	25.875240	25.334125	26.075651	25.683840	25.600423	64.387068

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on integer+100x1000 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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	colMeans2	119.515	133.136	138.7987	137.5605	143.5235	185.253
2	rowMeans2	144.974	165.378	172.8834	168.3720	175.8220	289.812

	expr	min	lq	mean	median	uq	max
1	colMeans2	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000
2	rowMeans2	1.213019	1.242173	1.245569	1.223985	1.22504	1.564412

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5234970 279.6    8529671 455.6  8529671 455.6
Vcells 9740368  74.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5234958 279.6    8529671 455.6  8529671 455.6
Vcells 9790401  74.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.081085	0.0905185	0.0962668	0.092616	0.0961065	0.151066
3	colMeans	0.085667	0.0928960	0.1003407	0.096000	0.1054990	0.151169
1	colMeans2	0.106836	0.1174530	0.1254939	0.122546	0.1274805	0.191988
4	apply+mean	0.903353	0.9894655	1.0771977	1.024349	1.1213500	1.578615

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000
3	colMeans	1.056509	1.026265	1.042319	1.036538	1.09773	1.000682
1	colMeans2	1.317580	1.297558	1.303605	1.323162	1.32645	1.270888
4	apply+mean	11.140815	10.931086	11.189713	11.060168	11.66779	10.449836

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.144888	0.153491	0.1650511	0.1635605	0.1693450	0.241371
2	.rowMeans	0.320223	0.346984	0.3659653	0.3583930	0.3687080	0.566216
3	rowMeans	0.322849	0.351780	0.3698612	0.3693370	0.3732825	0.496529
4	apply+mean	0.918096	1.004307	1.0543432	1.0237475	1.0430140	1.641317

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowMeans	2.210142	2.260615	2.217285	2.191195	2.177259	2.345833
3	rowMeans	2.228266	2.291861	2.240889	2.258106	2.204272	2.057120
4	apply+mean	6.336591	6.543103	6.387980	6.259137	6.159107	6.799976

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on integer+1000x100 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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	colMeans2	106.836	117.453	125.4939	122.5460	127.4805	191.988
2	rowMeans2	144.888	153.491	165.0511	163.5605	169.3450	241.371

	expr	min	lq	mean	median	uq	max
1	colMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMeans2	1.356172	1.306829	1.315212	1.334687	1.328399	1.257219

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5235197 279.6    8529671 455.6  8529671 455.6
Vcells 9856301  75.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5235182 279.6    8529671 455.6  8529671 455.6
Vcells 9856429  75.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	colMeans2	0.001996	0.0024525	0.0030822	0.0031190	0.0034125	0.015239
2	.colMeans	0.002793	0.0031900	0.0037191	0.0035125	0.0040765	0.014507
3	colMeans	0.005298	0.0057790	0.0066577	0.0063560	0.0073735	0.017426
4	apply+mean	0.065818	0.0689910	0.0720126	0.0704775	0.0726980	0.147648

	expr	min	lq	mean	median	uq	max
1	colMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.colMeans	1.399299	1.300714	1.206622	1.126162	1.194579	0.9519654
3	colMeans	2.654309	2.356371	2.160032	2.037833	2.160733	1.1435134
4	apply+mean	32.974950	28.130887	23.364032	22.596185	21.303443	9.6888247

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.002003	0.0022810	0.0029061	0.0029145	0.0032550	0.013025
2	.rowMeans	0.002305	0.0026760	0.0030477	0.0029510	0.0031760	0.007883
3	rowMeans	0.004476	0.0050145	0.0059769	0.0058160	0.0062755	0.031453
4	apply+mean	0.066499	0.0679560	0.0713963	0.0696305	0.0729915	0.139354

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.0000000	1.0000000
2	.rowMeans	1.150774	1.173170	1.048750	1.012524	0.9757296	0.6052207
3	rowMeans	2.234648	2.198378	2.056692	1.995540	1.9279570	2.4148177
4	apply+mean	33.199700	29.792196	24.568075	23.891062	22.4244240	10.6989635

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on double+10x10 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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	rowMeans2	2.003	2.2810	2.90606	2.9145	3.2550	13.025
1	colMeans2	1.996	2.4525	3.08220	3.1190	3.4125	15.239

	expr	min	lq	mean	median	uq	max
2	rowMeans2	1.0000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colMeans2	0.9965052	1.075186	1.060611	1.070166	1.048387	1.169981

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5235434 279.7    8529671 455.6  8529671 455.6
Vcells 9857359  75.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5235410 279.7    8529671 455.6  8529671 455.6
Vcells 9867372  75.3   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.008987	0.010441	0.0115130	0.011085	0.012182	0.018603
3	colMeans	0.011359	0.013062	0.0146494	0.013877	0.015752	0.034765
1	colMeans2	0.027815	0.030040	0.0336852	0.032915	0.034501	0.062917
4	apply+mean	0.410485	0.448217	0.5050557	0.488233	0.535023	0.757183

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colMeans	1.263937	1.251030	1.272422	1.251872	1.293055	1.868785
1	colMeans2	3.095026	2.877119	2.925835	2.969328	2.832129	3.382089
4	apply+mean	45.675420	42.928551	43.868258	44.044474	43.919143	40.702199

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.021014	0.0234945	0.0277081	0.0271525	0.031365	0.049027
2	.rowMeans	0.023541	0.0247930	0.0289273	0.0280110	0.030970	0.048662
3	rowMeans	0.024992	0.0281065	0.0323572	0.0316395	0.034588	0.051641
4	apply+mean	0.392591	0.4288000	0.4924598	0.4824895	0.520098	0.748713

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.0000000	1.0000000
2	.rowMeans	1.120253	1.055268	1.044001	1.031618	0.9874063	0.9925551
3	rowMeans	1.189302	1.196301	1.167789	1.165252	1.1027579	1.0533176
4	apply+mean	18.682355	18.251080	17.773162	17.769616	16.5821138	15.2714423

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on double+100x100 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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	rowMeans2	21.014	23.4945	27.70806	27.1525	31.365	49.027
1	colMeans2	27.815	30.0400	33.68517	32.9150	34.501	62.917

	expr	min	lq	mean	median	uq	max
2	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colMeans2	1.323641	1.278597	1.215717	1.212227	1.099984	1.283313

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5235640 279.7    8529671 455.6  8529671 455.6
Vcells 9858451  75.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5235634 279.7    8529671 455.6  8529671 455.6
Vcells 9868494  75.3   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.012308	0.0133250	0.0146995	0.0142910	0.0154300	0.028084
3	colMeans	0.014901	0.0161505	0.0176729	0.0172540	0.0186615	0.034493
1	colMeans2	0.031277	0.0332970	0.0369031	0.0363365	0.0400575	0.055325
4	apply+mean	0.152822	0.1561930	0.1708273	0.1665185	0.1801325	0.279692

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colMeans	1.210676	1.212045	1.202279	1.207333	1.209430	1.228208
1	colMeans2	2.541193	2.498837	2.510497	2.542614	2.596079	1.969983
4	apply+mean	12.416477	11.721801	11.621278	11.651984	11.674174	9.959123

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.024470	0.0260450	0.0282201	0.0273965	0.0296895	0.044917
2	.rowMeans	0.029456	0.0308975	0.0340446	0.0336540	0.0367775	0.043824
3	rowMeans	0.031758	0.0339380	0.0367688	0.0354640	0.0386305	0.054903
4	apply+mean	0.149294	0.1560205	0.1740137	0.1678455	0.1844835	0.288138

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.rowMeans	1.203760	1.186312	1.206397	1.228405	1.238738	0.9756662
3	rowMeans	1.297834	1.303052	1.302933	1.294472	1.301150	1.2223212
4	apply+mean	6.101103	5.990420	6.166313	6.126531	6.213762	6.4148986

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on double+1000x10 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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
2	rowMeans2	24.470	26.045	28.22006	27.3965	29.6895	44.917
1	colMeans2	31.277	33.297	36.90313	36.3365	40.0575	55.325

	expr	min	lq	mean	median	uq	max
2	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colMeans2	1.278177	1.278441	1.307691	1.326319	1.349214	1.231716

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5235880 279.7    8529671 455.6  8529671 455.6
Vcells 9858628  75.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5235856 279.7    8529671 455.6  8529671 455.6
Vcells 9868641  75.3   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.010296	0.0115655	0.0129187	0.0123925	0.0135930	0.022940
3	colMeans	0.012201	0.0142505	0.0166892	0.0154615	0.0176905	0.052161
1	colMeans2	0.029690	0.0341375	0.0367869	0.0356605	0.0379780	0.059502
4	apply+mean	3.155567	3.4715400	3.6900753	3.5772010	3.8181300	8.767973

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colMeans	1.185023	1.232156	1.291861	1.247650	1.301442	2.273801
1	colMeans2	2.883644	2.951667	2.847570	2.877587	2.793938	2.593810
4	apply+mean	306.484751	300.163417	285.638283	288.658544	280.889428	382.213296

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.024614	0.027924	0.0303011	0.0291305	0.0306365	0.070284
2	.rowMeans	0.025822	0.028817	0.0301902	0.0299545	0.0307390	0.049536
3	rowMeans	0.028587	0.031351	0.0335336	0.0323470	0.0342050	0.056989
4	apply+mean	3.134174	3.486367	3.6998181	3.5839590	3.8396975	8.790471

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	.rowMeans	1.049078	1.031980	0.9963397	1.028287	1.003346	0.7047977
3	rowMeans	1.161412	1.122726	1.1066797	1.110417	1.116479	0.8108389
4	apply+mean	127.332981	124.851991	122.1018541	123.031153	125.330815	125.0707273

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on double+10x1000 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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
2	rowMeans2	24.614	27.9240	30.30108	29.1305	30.6365	70.284
1	colMeans2	29.690	34.1375	36.78690	35.6605	37.9780	59.502

	expr	min	lq	mean	median	uq	max
2	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colMeans2	1.206224	1.222515	1.214046	1.224164	1.239632	0.8465938

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5236098 279.7    8529671 455.6  8529671 455.6
Vcells 9859933  75.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5236086 279.7    8529671 455.6  8529671 455.6
Vcells 9959966  76.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.066739	0.0727180	0.0806974	0.0767745	0.0853800	0.141699
3	colMeans	0.066169	0.0752170	0.0829405	0.0781310	0.0887805	0.126869
1	colMeans2	0.254845	0.2795845	0.2909625	0.2872260	0.2980720	0.369301
4	apply+mean	3.836750	4.3152365	4.6636250	4.4540265	4.6160400	21.206860

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.0000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	colMeans	0.9914593	1.034366	1.027797	1.017669	1.039828	0.8953415
1	colMeans2	3.8185319	3.844777	3.605601	3.741164	3.491122	2.6062358
4	apply+mean	57.4888746	59.342068	57.791537	58.014399	54.064652	149.6613244

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.198660	0.227482	0.2354485	0.2311605	0.2388610	0.320887
2	.rowMeans	0.212628	0.241880	0.2498996	0.2453545	0.2522545	0.370692
3	rowMeans	0.219002	0.239185	0.2494270	0.2483555	0.2561965	0.309427
4	apply+mean	3.825158	4.317397	4.6837063	4.4098160	4.6284035	21.396095

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.rowMeans	1.070311	1.063293	1.061377	1.061403	1.056072	1.1552104
3	rowMeans	1.102396	1.051446	1.059370	1.074385	1.072576	0.9642865
4	apply+mean	19.254797	18.979071	19.892699	19.076858	19.376974	66.6779739

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on double+100x1000 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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
2	rowMeans2	198.660	227.4820	235.4485	231.1605	238.861	320.887
1	colMeans2	254.845	279.5845	290.9625	287.2260	298.072	369.301

	expr	min	lq	mean	median	uq	max
2	rowMeans2	1.00000	1.00000	1.00000	1.000000	1.000000	1.000000
1	colMeans2	1.28282	1.22904	1.23578	1.242539	1.247889	1.150875

Figure: Benchmarking of colMeans2() and rowMeans2() 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 5236322 279.7    8529671 455.6  8529671 455.6
Vcells 9860082  75.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMeans2 = colMeans2(X, na.rm = FALSE), .colMeans = .colMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colMeans = colMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 2L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5236310 279.7    8529671 455.6  8529671 455.6
Vcells 9960115  76.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMeans2 = rowMeans2(X, na.rm = FALSE), .rowMeans = .rowMeans(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowMeans = rowMeans(X, na.rm = FALSE), `apply+mean` = apply(X, MARGIN = 1L, 
+     FUN = mean, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() 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	.colMeans	0.078356	0.0846185	0.0919227	0.0870375	0.0928480	0.152949
3	colMeans	0.081547	0.0872605	0.0946950	0.0910655	0.0991070	0.140971
1	colMeans2	0.234739	0.2545375	0.2689275	0.2617520	0.2790605	0.394196
4	apply+mean	1.028068	1.0913320	1.2594016	1.1235075	1.2650220	8.153290

	expr	min	lq	mean	median	uq	max
2	.colMeans	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	colMeans	1.040724	1.031222	1.030160	1.046279	1.067411	0.9216863
1	colMeans2	2.995801	3.008060	2.925584	3.007347	3.005563	2.5773035
4	apply+mean	13.120476	12.897085	13.700663	12.908315	13.624655	53.3072462

Table: Benchmarking of rowMeans2(), .rowMeans(), rowMeans() and apply+mean() 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	rowMeans2	0.188308	0.1979490	0.2157943	0.2110160	0.2248610	0.315616
2	.rowMeans	0.215012	0.2269885	0.2450801	0.2401625	0.2537825	0.375243
3	rowMeans	0.217828	0.2344560	0.2482889	0.2434065	0.2572280	0.350805
4	apply+mean	1.074836	1.1571190	1.3129454	1.1795160	1.2999040	8.436009

	expr	min	lq	mean	median	uq	max
1	rowMeans2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowMeans	1.141810	1.146702	1.135712	1.138125	1.128620	1.188923
3	rowMeans	1.156764	1.184426	1.150581	1.153498	1.143942	1.111493
4	apply+mean	5.707862	5.845541	6.084245	5.589699	5.780922	26.728711

Figure: Benchmarking of colMeans2(), .colMeans(), colMeans() and apply+mean() on double+1000x100 data as well as rowMeans2(), .rowMeans(), rowMeans() and apply+mean() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMeans2() and rowMeans2() 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
2	rowMeans2	188.308	197.9490	215.7943	211.016	224.8610	315.616
1	colMeans2	234.739	254.5375	268.9275	261.752	279.0605	394.196

	expr	min	lq	mean	median	uq	max
2	rowMeans2	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colMeans2	1.24657	1.285874	1.246221	1.240437	1.241036	1.248973

Figure: Benchmarking of colMeans2() and rowMeans2() 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 27.69 secs.

Reproducibility

To reproduce this report, do:

html <- matrixStats:::benchmark('colMeans2')

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 18:59:04 (+0200 UTC). Powered by RSP.