matrixStats.benchmarks

---

colMedians() and rowMedians() benchmarks

This report benchmark the performance of colMedians() and rowMedians() against alternative methods.

Alternative methods

• apply() + median()

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  5252365 280.6    7916910 422.9  7916910 422.9
Vcells 10194931  77.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5252172 280.5    7916910 422.9  7916910 422.9
Vcells 10194651  77.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() 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	colMedians	0.002242	0.0028040	0.0046619	0.0041435	0.0053185	0.017579
2	apply+median	0.301536	0.3285055	0.3668008	0.3471615	0.3814220	0.597683

	expr	min	lq	mean	median	uq	max
1	colMedians	1.0000	1.000	1.00000	1.0000	1.00000	1.00000
2	apply+median	134.4942	117.156	78.67969	83.7846	71.71609	33.99983

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	0.002414	0.0034080	0.0056612	0.0049215	0.0065285	0.017593
2	apply+median	0.311920	0.3353535	0.4024843	0.3581320	0.4088450	0.694812

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.0000	1.00000	1.00000	1.00000	1.00000	1.00000
2	apply+median	129.2129	98.40185	71.09523	72.76887	62.62465	39.49366

Figure: Benchmarking of colMedians() and apply+median() on integer+10x10 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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
1	colMedians	2.242	2.804	4.66195	4.1435	5.3185	17.579
2	rowMedians	2.414	3.408	5.66120	4.9215	6.5285	17.593

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians	1.076717	1.215407	1.214342	1.187764	1.227508	1.000796

Figure: Benchmarking of colMedians() and rowMedians() 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 5250734 280.5    7916910 422.9  7916910 422.9
Vcells 9811143  74.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5250728 280.5    7916910 422.9  7916910 422.9
Vcells 9816186  74.9   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on integer+100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	0.122332	0.132504	0.1460149	0.1434325	0.153560	0.280636
2	apply+median	2.631018	2.828682	3.2377973	3.0835700	3.355971	11.141453

	expr	min	lq	mean	median	uq	max
1	colMedians	1.00000	1.0000	1.00000	1.00000	1.00000	1.00000
2	apply+median	21.50719	21.3479	22.17443	21.49841	21.85446	39.70073

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	0.122275	0.125206	0.1336632	0.1298365	0.1347545	0.188559
2	apply+median	2.662727	2.726279	3.0154816	2.8214735	3.0111140	12.106425

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.00000	1.00000	1.0000	1.00000	1.00000	1.00000
2	apply+median	21.77654	21.77435	22.5603	21.73097	22.34518	64.20497

Figure: Benchmarking of colMedians() and apply+median() on integer+100x100 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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
2	rowMedians	122.275	125.206	133.6632	129.8365	134.7545	188.559
1	colMedians	122.332	132.504	146.0149	143.4325	153.5600	280.636

	expr	min	lq	mean	median	uq	max
2	rowMedians	1.000000	1.000000	1.00000	1.000000	1.000000	1.000000
1	colMedians	1.000466	1.058288	1.09241	1.104716	1.139554	1.488319

Figure: Benchmarking of colMedians() and rowMedians() 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 5251465 280.5    7916910 422.9  7916910 422.9
Vcells 9814650  74.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5251459 280.5    7916910 422.9  7916910 422.9
Vcells 9819693  75.0   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on integer+1000x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	0.109616	0.116921	0.1339929	0.1326660	0.143601	0.206983
2	apply+median	0.465915	0.496989	0.5680558	0.5575435	0.618175	0.923000

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.000000	1.00000	1.00000	1.000000
2	apply+median	4.250429	4.250639	4.239446	4.20261	4.30481	4.459303

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	0.111902	0.125378	0.1417642	0.140505	0.1542630	0.208260
2	apply+median	0.466662	0.521155	0.5982859	0.588065	0.6497965	0.937122

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.000000	1.00000	1.000000	1.000000	1.000000	1.00000
2	apply+median	4.170274	4.15667	4.220288	4.185367	4.212264	4.49977

Figure: Benchmarking of colMedians() and apply+median() on integer+1000x10 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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	colMedians	109.616	116.921	133.9929	132.666	143.601	206.983
2	rowMedians	111.902	125.378	141.7643	140.505	154.263	208.260

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
2	rowMedians	1.020855	1.072331	1.057998	1.059088	1.074247	1.00617

Figure: Benchmarking of colMedians() and rowMedians() 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 5251653 280.5    7916910 422.9  7916910 422.9
Vcells 9815325  74.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5251647 280.5    7916910 422.9  7916910 422.9
Vcells 9820368  75.0   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on integer+10x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	0.131982	0.1396325	0.159328	0.1586865	0.164361	0.651198
2	apply+median	24.277239	24.8847200	26.678439	25.8006730	26.610819	37.838270

	expr	min	lq	mean	median	uq	max
1	colMedians	1.0000	1.0000	1.0000	1.000	1.0000	1.00000
2	apply+median	183.9436	178.2158	167.4435	162.589	161.9047	58.10563

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	0.132675	0.1409135	0.1565327	0.1614335	0.1688645	0.198457
2	apply+median	24.261659	25.1337500	26.6236012	25.6448825	26.3032430	34.410864

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.0000	1.000	1.0000	1.0000	1.0000	1.000
2	apply+median	182.8653	178.363	170.0833	158.8573	155.7654	173.392

Figure: Benchmarking of colMedians() and apply+median() on integer+10x1000 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

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

	expr	min	lq	mean	median	uq	max
1	colMedians	131.982	139.6325	159.3280	158.6865	164.3610	651.198
2	rowMedians	132.675	140.9135	156.5327	161.4335	168.8645	198.457

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.0000000	1.000000	1.0000	1.0000000
2	rowMedians	1.005251	1.009174	0.9824554	1.017311	1.0274	0.3047568

Figure: Benchmarking of colMedians() and rowMedians() 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 5251834 280.5    7916910 422.9  7916910 422.9
Vcells 9815806  74.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5251828 280.5    7916910 422.9  7916910 422.9
Vcells 9865849  75.3   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on integer+100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	1.211458	1.250171	1.296478	1.288948	1.316439	1.506328
2	apply+median	26.522826	27.794670	33.481847	28.772193	29.446193	394.031374

	expr	min	lq	mean	median	uq	max
1	colMedians	1.00000	1.00000	1.00000	1.00000	1.00000	1.000
2	apply+median	21.89331	22.23269	25.82524	22.32223	22.36807	261.584

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	1.221957	1.288993	1.34172	1.339344	1.371976	1.596619
2	apply+median	26.951488	28.553535	30.57973	29.152181	29.917817	43.257642

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.000	1.00000	1.00000	1.00000	1.00000	1.00000
2	apply+median	22.056	22.15181	22.79143	21.76602	21.80638	27.09328

Figure: Benchmarking of colMedians() and apply+median() on integer+100x1000 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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	colMedians	1.211458	1.250171	1.296478	1.288948	1.316439	1.506328
2	rowMedians	1.221957	1.288993	1.341720	1.339344	1.371976	1.596619

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians	1.008666	1.031054	1.034896	1.039098	1.042187	1.059941

Figure: Benchmarking of colMedians() and rowMedians() 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 5252030 280.5    7916910 422.9  7916910 422.9
Vcells 9816385  74.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5252024 280.5    7916910 422.9  7916910 422.9
Vcells 9866428  75.3   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on integer+1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	1.060225	1.116370	1.176325	1.125428	1.163722	1.755365
2	apply+median	4.404879	4.618464	5.186349	4.710299	4.974653	15.663975

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+median	4.154664	4.137035	4.408942	4.185338	4.274776	8.923486

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	1.075615	1.086423	1.146017	1.136283	1.157306	1.512098
2	apply+median	4.367497	4.561608	4.863605	4.663398	4.738879	14.860419

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+median	4.060465	4.198742	4.243921	4.104083	4.094752	9.827682

Figure: Benchmarking of colMedians() and apply+median() on integer+1000x100 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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	colMedians	1.060225	1.116370	1.176325	1.125428	1.163722	1.755365
2	rowMedians	1.075615	1.086423	1.146017	1.136283	1.157306	1.512098

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.0000000	1.0000000	1.000000	1.0000000	1.0000000
2	rowMedians	1.014516	0.9731738	0.9742349	1.009644	0.9944858	0.8614151

Figure: Benchmarking of colMedians() and rowMedians() 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 5252236 280.5    7916910 422.9  7916910 422.9
Vcells 9932799  75.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5252215 280.5    7916910 422.9  7916910 422.9
Vcells 9932917  75.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() 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	colMedians	0.003102	0.0037710	0.0052201	0.0047445	0.006130	0.019869
2	apply+median	0.307751	0.3289145	0.3585106	0.3435010	0.375892	0.648243

	expr	min	lq	mean	median	uq	max
1	colMedians	1.00000	1.00000	1.00000	1.00000	1.00000	1.00000
2	apply+median	99.21051	87.22209	68.67874	72.39983	61.32007	32.62585

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	0.003381	0.0042770	0.0058312	0.0059115	0.0066925	0.019572
2	apply+median	0.325495	0.3390405	0.3764986	0.3631345	0.4141270	0.617893

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.00000	1.00000	1.00000	1.00000	1.00000	1.00000
2	apply+median	96.27181	79.27063	64.56568	61.42849	61.87927	31.57025

Figure: Benchmarking of colMedians() and apply+median() on double+10x10 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

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

	expr	min	lq	mean	median	uq	max
1	colMedians	3.102	3.771	5.22011	4.7445	6.1300	19.869
2	rowMedians	3.381	4.277	5.83125	5.9115	6.6925	19.572

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowMedians	1.089942	1.134182	1.117074	1.245969	1.091762	0.9850521

Figure: Benchmarking of colMedians() and rowMedians() 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 5252408 280.6    7916910 422.9  7916910 422.9
Vcells 9932899  75.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5252402 280.6    7916910 422.9  7916910 422.9
Vcells 9942942  75.9   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on double+100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	0.182377	0.1866275	0.2042947	0.191434	0.2238895	0.33058
2	apply+median	2.738506	2.7896065	3.1864199	2.937558	3.2481650	12.31540

	expr	min	lq	mean	median	uq	max
1	colMedians	1.00000	1.00000	1.00000	1.00000	1.00000	1.00000
2	apply+median	15.01563	14.94746	15.59717	15.34501	14.50789	37.25391

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	0.182541	0.1863345	0.2039631	0.192911	0.220481	0.27028
2	apply+median	2.741946	2.7972355	3.1691965	2.912469	3.183548	12.31023

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.00000	1.0000	1.00000	1.00000	1.0000	1.00000
2	apply+median	15.02099	15.0119	15.53809	15.09747	14.4391	45.54621

Figure: Benchmarking of colMedians() and apply+median() on double+100x100 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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	colMedians	182.377	186.6275	204.2947	191.434	223.8895	330.58
2	rowMedians	182.541	186.3345	203.9631	192.911	220.4810	270.28

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.00000	1.0000000	1.000000	1.000000	1.0000000
2	rowMedians	1.000899	0.99843	0.9983767	1.007715	0.984776	0.8175933

Figure: Benchmarking of colMedians() and rowMedians() 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 5252599 280.6    7916910 422.9  7916910 422.9
Vcells 9933796  75.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5252593 280.6    7916910 422.9  7916910 422.9
Vcells 9943839  75.9   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on double+1000x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	0.169833	0.1745720	0.2093347	0.202332	0.2230540	0.316110
2	apply+median	0.528834	0.5411475	0.6521944	0.592137	0.7022785	1.435467

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+median	3.113847	3.099853	3.115558	2.926561	3.148468	4.541036

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	0.170217	0.1801315	0.2080748	0.1969065	0.2254305	0.302917
2	apply+median	0.527815	0.5518570	0.6374527	0.6040875	0.6901700	1.032708

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000
2	apply+median	3.100836	3.063634	3.063575	3.06789	3.061564	3.409211

Figure: Benchmarking of colMedians() and apply+median() on double+1000x10 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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	rowMedians	170.217	180.1315	208.0748	196.9065	225.4305	302.917
1	colMedians	169.833	174.5720	209.3347	202.3320	223.0540	316.110

	expr	min	lq	mean	median	uq	max
2	rowMedians	1.0000000	1.0000000	1.000000	1.000000	1.0000000	1.000000
1	colMedians	0.9977441	0.9691364	1.006055	1.027554	0.9894579	1.043553

Figure: Benchmarking of colMedians() and rowMedians() 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 5252787 280.6    7916910 422.9  7916910 422.9
Vcells 9934842  75.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5252781 280.6    7916910 422.9  7916910 422.9
Vcells 9944885  75.9   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on double+10x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	0.192799	0.199092	0.2134576	0.211814	0.2179785	0.305906
2	apply+median	24.453860	25.212035	26.7656468	25.614391	26.2556555	36.038124

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000	1.0000	1.0000	1.0000	1.0000	1.0000
2	apply+median	126.836	126.6351	125.3909	120.9287	120.4507	117.8078

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	0.19287	0.199287	0.2125211	0.2126835	0.218602	0.275648
2	apply+median	24.39521	25.258805	26.8713863	25.9733350	26.334249	36.129646

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.0000	1.0000	1.000	1.000	1.0000	1.0000
2	apply+median	126.4853	126.7459	126.441	122.122	120.4666	131.0717

Figure: Benchmarking of colMedians() and apply+median() on double+10x1000 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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	colMedians	192.799	199.092	213.4576	211.8140	217.9785	305.906
2	rowMedians	192.870	199.287	212.5211	212.6835	218.6020	275.648

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.0000000	1.000000	1.00000	1.0000000
2	rowMedians	1.000368	1.000979	0.9956129	1.004105	1.00286	0.9010873

Figure: Benchmarking of colMedians() and rowMedians() 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 5252968 280.6    7916910 422.9  7916910 422.9
Vcells 9934960  75.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5252962 280.6    7916910 422.9  7916910 422.9
Vcells 10035003  76.6   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on double+100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	1.788286	1.847881	1.90678	1.884424	1.92834	2.272694
2	apply+median	27.450899	28.380057	30.94689	29.288891	30.05219	46.750083

	expr	min	lq	mean	median	uq	max
1	colMedians	1.0000	1.00000	1.00000	1.00000	1.00000	1.00000
2	apply+median	15.3504	15.35816	16.22991	15.54262	15.58449	20.57034

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	1.830289	1.928441	1.995154	1.979102	2.042293	2.365565
2	apply+median	27.513177	28.842820	31.300535	29.459474	30.431528	57.828117

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.00000	1.00000	1.00000	1.00000	1.00000	1.0000
2	apply+median	15.03215	14.95655	15.68828	14.88527	14.90067	24.4458

Figure: Benchmarking of colMedians() and apply+median() on double+100x1000 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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	colMedians	1.788286	1.847881	1.906780	1.884424	1.928340	2.272694
2	rowMedians	1.830289	1.928441	1.995154	1.979102	2.042293	2.365565

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians	1.023488	1.043596	1.046347	1.050242	1.059094	1.040864

Figure: Benchmarking of colMedians() and rowMedians() 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 5253164 280.6    7916910 422.9  7916910 422.9
Vcells 9936196  75.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians = colMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 2L, 
+     FUN = median, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5253158 280.6    7916910 422.9  7916910 422.9
Vcells 10036239  76.6   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians = rowMedians(X, na.rm = FALSE), `apply+median` = apply(X, MARGIN = 1L, 
+     FUN = median, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colMedians() and apply+median() on double+1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colMedians	1.669742	1.711053	1.807156	1.755450	1.793972	2.815782
2	apply+median	5.022684	5.148174	5.525504	5.247178	5.320824	15.706835

	expr	min	lq	mean	median	uq	max
1	colMedians	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+median	3.00806	3.008776	3.057569	2.989079	2.965947	5.578143

Table: Benchmarking of rowMedians() and apply+median() 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	rowMedians	1.687571	1.730504	1.819095	1.761198	1.807409	3.171408
2	apply+median	5.078287	5.193276	5.703436	5.274445	5.370983	14.610601

	expr	min	lq	mean	median	uq	max
1	rowMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+median	3.009229	3.001019	3.135315	2.994804	2.971647	4.606976

Figure: Benchmarking of colMedians() and apply+median() on double+1000x100 data as well as rowMedians() and apply+median() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMedians() and rowMedians() 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	colMedians	1.669742	1.711053	1.807156	1.755450	1.793972	2.815782
2	rowMedians	1.687571	1.730504	1.819095	1.761198	1.807409	3.171408

	expr	min	lq	mean	median	uq	max
1	colMedians	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians	1.010678	1.011368	1.006607	1.003275	1.007491	1.126297

Figure: Benchmarking of colMedians() and rowMedians() on double+1000x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

Appendix

Session information

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

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

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

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

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

loaded via a namespace (and not attached):
 [1] Biobase_2.52.0          httr_1.4.2              splines_4.1.1          
 [4] bit64_4.0.5             network_1.17.1          assertthat_0.2.1       
 [7] highr_0.9               stats4_4.1.1            blob_1.2.2             
[10] GenomeInfoDbData_1.2.6  robustbase_0.93-8       pillar_1.6.2           
[13] RSQLite_2.2.8           lattice_0.20-44         glue_1.4.2             
[16] digest_0.6.27           XVector_0.32.0          colorspace_2.0-2       
[19] Matrix_1.3-4            XML_3.99-0.7            pkgconfig_2.0.3        
[22] zlibbioc_1.38.0         genefilter_1.74.0       purrr_0.3.4            
[25] ergm_4.1.2              xtable_1.8-4            scales_1.1.1           
[28] tibble_3.1.4            annotate_1.70.0         KEGGREST_1.32.0        
[31] farver_2.1.0            generics_0.1.0          IRanges_2.26.0         
[34] ellipsis_0.3.2          cachem_1.0.6            withr_2.4.2            
[37] BiocGenerics_0.38.0     mime_0.11               survival_3.2-13        
[40] magrittr_2.0.1          crayon_1.4.1            statnet.common_4.5.0   
[43] memoise_2.0.0           laeken_0.5.1            fansi_0.5.0            
[46] R.cache_0.15.0          MASS_7.3-54             R.rsp_0.44.0           
[49] progressr_0.8.0         tools_4.1.1             lifecycle_1.0.0        
[52] S4Vectors_0.30.0        trust_0.1-8             munsell_0.5.0          
[55] tabby_0.0.1-9001        AnnotationDbi_1.54.1    Biostrings_2.60.2      
[58] compiler_4.1.1          GenomeInfoDb_1.28.1     rlang_0.4.11           
[61] grid_4.1.1              RCurl_1.98-1.4          cwhmisc_6.6            
[64] rstudioapi_0.13         rappdirs_0.3.3          startup_0.15.0         
[67] labeling_0.4.2          bitops_1.0-7            base64enc_0.1-3        
[70] boot_1.3-28             gtable_0.3.0            DBI_1.1.1              
[73] markdown_1.1            R6_2.5.1                lpSolveAPI_5.5.2.0-17.7
[76] rle_0.9.2               dplyr_1.0.7             fastmap_1.1.0          
[79] bit_4.0.4               utf8_1.2.2              parallel_4.1.1         
[82] Rcpp_1.0.7              vctrs_0.3.8             png_0.1-7              
[85] DEoptimR_1.0-9          tidyselect_1.1.1        xfun_0.25              
[88] coda_0.19-4            

Total processing time was 51.72 secs.

Reproducibility

To reproduce this report, do:

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

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 22:20:57 (+0200 UTC). Powered by RSP.