matrixStats.benchmarks

---

colMins() and rowMins() benchmarks

This report benchmark the performance of colMins() and rowMins() against alternative methods.

Alternative methods

• apply() + min()
• lapply() + pmin()
• lapply() + pmin.int()

See also StackOverflow:colMins?.

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  5261045 281.0    8529671 455.6  8529671 455.6
Vcells 10261516  78.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5250822 280.5    8529671 455.6  8529671 455.6
Vcells 10228244  78.1   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.001961	0.0024625	0.0031184	0.0030480	0.0035400	0.013601
4	lapply+pmin.int	0.019177	0.0207965	0.0218450	0.0213775	0.0220245	0.058372
3	lapply+pmin	0.030621	0.0336390	0.0348887	0.0345025	0.0353340	0.061177
2	apply+min	0.032676	0.0351865	0.0383533	0.0376110	0.0396480	0.092919

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	9.779194	8.445279	7.005259	7.013616	6.221610	4.291743
3	lapply+pmin	15.614992	13.660508	11.188108	11.319718	9.981356	4.497978
2	apply+min	16.662927	14.288934	12.299140	12.339567	11.200000	6.831777

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.002075	0.0025490	0.0032280	0.003185	0.003531	0.015027
4	lapply+pmin.int	0.018828	0.0199445	0.0209203	0.020780	0.021589	0.030489
3	lapply+pmin	0.030842	0.0327230	0.0345111	0.034101	0.034914	0.076136
2	apply+min	0.032612	0.0361775	0.0386453	0.037367	0.039665	0.088746

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	9.073735	7.824441	6.480926	6.524333	6.114132	2.028948
3	lapply+pmin	14.863615	12.837583	10.691250	10.706750	9.887851	5.066613
2	apply+min	15.716627	14.192821	11.971979	11.732182	11.233362	5.905770

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on integer+10x10 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	1.961	2.4625	3.11837	3.048	3.540	13.601
2	rowMins	2.075	2.5490	3.22798	3.185	3.531	15.027

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.00000	1.000000	1.0000000	1.000000
2	rowMins	1.058134	1.035127	1.03515	1.044947	0.9974576	1.104845

Figure: Benchmarking of colMins() and rowMins() 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 5249419 280.4    8529671 455.6  8529671 455.6
Vcells 9845021  75.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5249395 280.4    8529671 455.6  8529671 455.6
Vcells 9850098  75.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.019701	0.0220015	0.0260000	0.0243420	0.0275535	0.050158
4	lapply+pmin.int	0.160862	0.1737545	0.2046270	0.1904940	0.2215825	0.342907
2	apply+min	0.162650	0.1761235	0.2063965	0.1980970	0.2242105	0.357808
3	lapply+pmin	0.218463	0.2366050	0.2765518	0.2609345	0.2945850	0.506499

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	8.165169	7.897393	7.870271	7.825733	8.041900	6.836537
2	apply+min	8.255926	8.005068	7.938332	8.138074	8.137278	7.133618
3	lapply+pmin	11.088930	10.754039	10.636612	10.719518	10.691382	10.098070

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.028062	0.0313810	0.0370879	0.0366945	0.0410420	0.059481
4	lapply+pmin.int	0.140107	0.1504120	0.1716608	0.1668850	0.1818020	0.293210
2	apply+min	0.161842	0.1730955	0.2021828	0.1897060	0.2156300	0.380225
3	lapply+pmin	0.191663	0.2107970	0.2371760	0.2327130	0.2503695	0.338557

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	4.992766	4.793091	4.628482	4.547957	4.429657	4.929473
2	apply+min	5.767301	5.515933	5.451446	5.169876	5.253886	6.392377
3	lapply+pmin	6.829984	6.717345	6.394967	6.341904	6.100324	5.691851

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on integer+100x100 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	19.701	22.0015	25.99999	24.3420	27.5535	50.158
2	rowMins	28.062	31.3810	37.08792	36.6945	41.0420	59.481

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins	1.424395	1.426312	1.426459	1.507456	1.489539	1.185873

Figure: Benchmarking of colMins() and rowMins() 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 5250165 280.4    8529671 455.6  8529671 455.6
Vcells 9848772  75.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5250159 280.4    8529671 455.6  8529671 455.6
Vcells 9853879  75.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.016098	0.0182495	0.0195772	0.0189285	0.0201150	0.034340
2	apply+min	0.092033	0.0980675	0.1057655	0.1039540	0.1100675	0.178412
4	lapply+pmin.int	1.041009	1.1513665	1.2030655	1.1846995	1.2500025	1.630197
3	lapply+pmin	1.553083	1.6863190	1.8311417	1.7646325	1.8476255	7.274179

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000
2	apply+min	5.717046	5.373709	5.402473	5.49193	5.471912	5.195457
4	lapply+pmin.int	64.666977	63.090304	61.452255	62.58813	62.142804	47.472248
3	lapply+pmin	96.476767	92.403573	93.534212	93.22622	91.853120	211.828160

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.027722	0.0322975	0.0355971	0.0339830	0.035857	0.070898
2	apply+min	0.091952	0.1015750	0.1117716	0.1064265	0.117356	0.235355
4	lapply+pmin.int	1.022871	1.1123160	1.2268829	1.1674330	1.210100	6.624442
3	lapply+pmin	1.516378	1.6735670	1.7703788	1.7543370	1.824891	2.640951

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
2	apply+min	3.316932	3.14498	3.139912	3.131757	3.272889	3.319628
4	lapply+pmin.int	36.897446	34.43969	34.465849	34.353441	33.747929	93.436232
3	lapply+pmin	54.699444	51.81723	49.733848	51.623959	50.893577	37.250007

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on integer+1000x10 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	16.098	18.2495	19.57724	18.9285	20.115	34.340
2	rowMins	27.722	32.2975	35.59706	33.9830	35.857	70.898

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.0000	1.000000
2	rowMins	1.722077	1.769775	1.818288	1.795335	1.7826	2.064589

Figure: Benchmarking of colMins() and rowMins() 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 5250393 280.5    8529671 455.6  8529671 455.6
Vcells 9849681  75.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5250381 280.5    8529671 455.6  8529671 455.6
Vcells 9854778  75.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.040626	0.0447915	0.0499942	0.0475190	0.0517620	0.078766
4	lapply+pmin.int	0.086942	0.0944890	0.1020910	0.0979325	0.1051145	0.148585
3	lapply+pmin	0.096333	0.1060625	0.1173308	0.1099185	0.1234000	0.203392
2	apply+min	0.938481	1.0313715	1.0971496	1.0732180	1.1306615	1.537324

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	2.140058	2.109530	2.042059	2.060913	2.030727	1.886410
3	lapply+pmin	2.371215	2.367916	2.346891	2.313148	2.383988	2.582231
2	apply+min	23.100502	23.026054	21.945560	22.585029	21.843466	19.517609

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.040354	0.0436515	0.0481795	0.0461620	0.0510930	0.087116
4	lapply+pmin.int	0.067021	0.0711615	0.0793412	0.0754335	0.0841045	0.138087
3	lapply+pmin	0.075066	0.0816025	0.0896339	0.0858935	0.0951510	0.128482
2	apply+min	0.935962	1.0378985	1.0887489	1.0698970	1.1208725	1.622883

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	1.660827	1.630219	1.646785	1.634104	1.646106	1.585093
3	lapply+pmin	1.860187	1.869409	1.860417	1.860697	1.862310	1.474838
2	apply+min	23.193785	23.776926	22.597777	23.177007	21.937888	18.628989

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on integer+10x1000 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	rowMins	40.354	43.6515	48.17947	46.162	51.093	87.116
1	colMins	40.626	44.7915	49.99415	47.519	51.762	78.766

	expr	min	lq	mean	median	uq	max
2	rowMins	1.00000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colMins	1.00674	1.026116	1.037665	1.029397	1.013094	0.9041508

Figure: Benchmarking of colMins() and rowMins() 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 5250623 280.5    8529671 455.6  8529671 455.6
Vcells 9850292  75.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5250611 280.5    8529671 455.6  8529671 455.6
Vcells 9900389  75.6   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.185483	0.212613	0.2210963	0.2201230	0.2279455	0.292429
4	lapply+pmin.int	0.707608	0.800243	0.8401222	0.8207525	0.8510785	1.202391
3	lapply+pmin	0.765025	0.855969	0.8861484	0.8817275	0.8931765	1.216636
2	apply+min	1.578081	1.709996	1.9386553	1.7540850	1.7909110	17.400191

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	3.814948	3.763848	3.799803	3.728609	3.733693	4.111737
3	lapply+pmin	4.124502	4.025949	4.007975	4.005613	3.918377	4.160449
2	apply+min	8.507955	8.042765	8.768377	7.968658	7.856751	59.502276

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.272467	0.2965420	0.3111866	0.3066340	0.3163490	0.408973
4	lapply+pmin.int	0.488489	0.5373605	0.5648683	0.5483525	0.5768830	0.789271
3	lapply+pmin	0.543294	0.6062425	0.6287185	0.6221345	0.6356095	0.819154
2	apply+min	1.541115	1.6870285	1.9138076	1.7120345	1.7813550	17.143776

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	1.792837	1.812089	1.815208	1.788296	1.823565	1.929885
3	lapply+pmin	1.993981	2.044373	2.020391	2.028916	2.009203	2.002954
2	apply+min	5.656153	5.689004	6.150032	5.583316	5.630980	41.919090

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on integer+100x1000 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	185.483	212.613	221.0963	220.123	227.9455	292.429
2	rowMins	272.467	296.542	311.1866	306.634	316.3490	408.973

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
2	rowMins	1.468959	1.39475	1.407471	1.393012	1.387827	1.398538

Figure: Benchmarking of colMins() and rowMins() 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 5250841 280.5    8529671 455.6  8529671 455.6
Vcells 9851077  75.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5250835 280.5    8529671 455.6  8529671 455.6
Vcells 9901184  75.6   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.142558	0.1614985	0.1669243	0.1647525	0.1715450	0.217975
2	apply+min	0.710634	0.7904130	0.8290255	0.8108770	0.8435595	1.149814
4	lapply+pmin.int	1.654945	1.8326940	2.1036546	1.8882600	1.9360495	21.940809
3	lapply+pmin	2.126190	2.4123750	2.4756398	2.4662435	2.5552180	3.071616

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+min	4.984876	4.894244	4.966477	4.921789	4.917424	5.274981
4	lapply+pmin.int	11.608924	11.348056	12.602448	11.461192	11.285957	100.657456
3	lapply+pmin	14.914561	14.937445	14.830915	14.969384	14.895322	14.091598

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.262699	0.2928810	0.3032057	0.2963325	0.3085615	0.397681
2	apply+min	0.723057	0.7841425	0.8252375	0.7958415	0.8274980	1.205860
4	lapply+pmin.int	1.407189	1.5698125	1.9452756	1.6088105	1.6758330	32.196058
3	lapply+pmin	1.885795	2.1447950	2.2211179	2.2088455	2.2686420	2.785102

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+min	2.752416	2.677342	2.721708	2.685637	2.681793	3.032229
4	lapply+pmin.int	5.356659	5.359899	6.415695	5.429072	5.431115	80.959508
3	lapply+pmin	7.178539	7.323094	7.325448	7.453943	7.352317	7.003357

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on integer+1000x100 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	142.558	161.4985	166.9243	164.7525	171.5450	217.975
2	rowMins	262.699	292.8810	303.2057	296.3325	308.5615	397.681

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000
2	rowMins	1.842752	1.813522	1.816427	1.798652	1.79872	1.824434

Figure: Benchmarking of colMins() and rowMins() 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 5251084 280.5    8529671 455.6  8529671 455.6
Vcells 9967044  76.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5251057 280.5    8529671 455.6  8529671 455.6
Vcells 9967216  76.1   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.001998	0.0025055	0.0031965	0.0031380	0.0034950	0.013934
4	lapply+pmin.int	0.019558	0.0210665	0.0221236	0.0215675	0.0225060	0.052832
3	lapply+pmin	0.031477	0.0339570	0.0349273	0.0345505	0.0354165	0.052220
2	apply+min	0.033463	0.0356130	0.0384171	0.0377820	0.0395280	0.094447

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	9.788789	8.408102	6.921167	6.873008	6.439485	3.791589
3	lapply+pmin	15.754254	13.552983	10.926692	11.010357	10.133476	3.747668
2	apply+min	16.748248	14.213929	12.018467	12.040153	11.309871	6.778168

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.001981	0.0023840	0.0031251	0.0030920	0.0035575	0.014820
4	lapply+pmin.int	0.018385	0.0199360	0.0209329	0.0208635	0.0216685	0.030255
3	lapply+pmin	0.029550	0.0325610	0.0343982	0.0338975	0.0351370	0.074499
2	apply+min	0.032283	0.0359755	0.0385445	0.0373450	0.0393505	0.088373

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	9.280666	8.362416	6.698369	6.747574	6.090935	2.041498
3	lapply+pmin	14.916709	13.658138	11.007129	10.962969	9.876880	5.026923
2	apply+min	16.296315	15.090394	12.333918	12.077943	11.061279	5.963090

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on double+10x10 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	rowMins	1.981	2.3840	3.12508	3.092	3.5575	14.820
1	colMins	1.998	2.5055	3.19651	3.138	3.4950	13.934

	expr	min	lq	mean	median	uq	max
2	rowMins	1.000000	1.000000	1.000000	1.000000	1.0000000	1.0000000
1	colMins	1.008582	1.050965	1.022857	1.014877	0.9824315	0.9402159

Figure: Benchmarking of colMins() and rowMins() 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 5251309 280.5    8529671 455.6  8529671 455.6
Vcells 9968120  76.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5251285 280.5    8529671 455.6  8529671 455.6
Vcells 9978197  76.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.023677	0.0261010	0.0301559	0.0291890	0.0317655	0.052084
2	apply+min	0.165868	0.1785005	0.2056516	0.1978490	0.2189710	0.363973
4	lapply+pmin.int	0.172886	0.1855755	0.2154075	0.2046485	0.2312620	0.433534
3	lapply+pmin	0.228570	0.2484515	0.2819802	0.2740650	0.2975335	0.459524

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+min	7.005448	6.838838	6.819611	6.778204	6.893359	6.988192
4	lapply+pmin.int	7.301854	7.109900	7.143128	7.011152	7.280288	8.323746
3	lapply+pmin	9.653672	9.518850	9.350746	9.389325	9.366561	8.822748

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.026255	0.0296210	0.0347608	0.0339850	0.0384645	0.055810
4	lapply+pmin.int	0.147401	0.1611060	0.1827663	0.1819535	0.1937765	0.298389
2	apply+min	0.167141	0.1795820	0.2087781	0.2038960	0.2210330	0.380344
3	lapply+pmin	0.201589	0.2220815	0.2505038	0.2475970	0.2627900	0.375251

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	5.614207	5.438912	5.257834	5.353936	5.037801	5.346515
2	apply+min	6.366064	6.062658	6.006142	5.999588	5.746416	6.814979
3	lapply+pmin	7.678119	7.497434	7.206509	7.285479	6.832014	6.723723

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on double+100x100 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	23.677	26.101	30.15591	29.189	31.7655	52.084
2	rowMins	26.255	29.621	34.76077	33.985	38.4645	55.810

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins	1.108882	1.134861	1.152702	1.164308	1.210889	1.071538

Figure: Benchmarking of colMins() and rowMins() 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 5251515 280.5    8529671 455.6  8529671 455.6
Vcells 9969241  76.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5251509 280.5    8529671 455.6  8529671 455.6
Vcells 9979348  76.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.021997	0.0242860	0.0265977	0.025224	0.0265595	0.068963
2	apply+min	0.096774	0.1063215	0.1158731	0.110821	0.1172005	0.235033
4	lapply+pmin.int	1.038897	1.1553585	1.2204119	1.188010	1.2416395	2.431739
3	lapply+pmin	1.530459	1.7143695	1.8709654	1.766916	1.8631005	6.630570

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+min	4.399418	4.377893	4.356508	4.393474	4.412752	3.408103
4	lapply+pmin.int	47.229031	47.573026	45.884115	47.098398	46.749355	35.261502
3	lapply+pmin	69.575806	70.590855	70.343125	70.048981	70.148177	96.146774

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.023627	0.0271900	0.0292032	0.0287020	0.0300185	0.052191
2	apply+min	0.095888	0.1071055	0.1161132	0.1125895	0.1192445	0.192410
4	lapply+pmin.int	1.017413	1.1128380	1.2169178	1.1503685	1.1986465	6.657355
3	lapply+pmin	1.525861	1.6821610	1.7586292	1.7375345	1.8250130	2.348565

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
2	apply+min	4.058408	3.93915	3.976042	3.922706	3.972367	3.686651
4	lapply+pmin.int	43.061455	40.92821	41.670673	40.079733	39.930260	127.557529
3	lapply+pmin	64.581242	61.86690	60.220387	60.537053	60.796276	44.999425

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on double+1000x10 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	21.997	24.286	26.59770	25.224	26.5595	68.963
2	rowMins	23.627	27.190	29.20322	28.702	30.0185	52.191

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.00000	1.000000	1.000000	1.0000000
2	rowMins	1.074101	1.119575	1.09796	1.137885	1.130236	0.7567971

Figure: Benchmarking of colMins() and rowMins() 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 5251743 280.5    8529671 455.6  8529671 455.6
Vcells 9969398  76.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5251731 280.5    8529671 455.6  8529671 455.6
Vcells 9979495  76.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.039752	0.0437190	0.0484646	0.0462090	0.0504145	0.079809
4	lapply+pmin.int	0.100573	0.1097125	0.1185558	0.1143175	0.1255090	0.167419
3	lapply+pmin	0.109426	0.1209245	0.1339786	0.1264700	0.1412060	0.229594
2	apply+min	0.938959	1.0064870	1.0852165	1.0653735	1.1191365	1.535932

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	2.530011	2.509492	2.446234	2.473923	2.489542	2.097746
3	lapply+pmin	2.752717	2.765948	2.764462	2.736913	2.800901	2.876793
2	apply+min	23.620422	23.021730	22.391937	23.055541	22.198703	19.245098

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.040033	0.0433390	0.0484441	0.046561	0.0517350	0.091275
4	lapply+pmin.int	0.071748	0.0770780	0.0892832	0.083807	0.0964775	0.141539
3	lapply+pmin	0.081649	0.0894985	0.0988413	0.094670	0.1046090	0.151546
2	apply+min	0.936921	1.0311710	1.0894527	1.064358	1.1025045	1.489842

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	1.792221	1.77849	1.843014	1.799940	1.864840	1.550687
3	lapply+pmin	2.039542	2.06508	2.040316	2.033247	2.022016	1.660323
2	apply+min	23.403717	23.79314	22.488858	22.859432	21.310612	16.322564

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on double+10x1000 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	39.752	43.719	48.46461	46.209	50.4145	79.809
2	rowMins	40.033	43.339	48.44411	46.561	51.7350	91.275

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.0000000	1.000000	1.000000	1.000000	1.000000
2	rowMins	1.007069	0.9913081	0.999577	1.007618	1.026193	1.143668

Figure: Benchmarking of colMins() and rowMins() 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 5251973 280.5    8529671 455.6  8529671 455.6
Vcells 9970756  76.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5251961 280.5    8529671 455.6  8529671 455.6
Vcells 10070853  76.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.227381	0.2548190	0.2739132	0.267334	0.277895	0.502656
4	lapply+pmin.int	0.863405	0.9637765	1.1428165	0.987881	1.027360	14.176907
3	lapply+pmin	0.915867	0.9997155	1.0604782	1.050261	1.085116	1.476039
2	apply+min	1.584813	1.6895700	1.9285642	1.748488	1.850658	14.398383

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	3.797173	3.782200	4.172184	3.695306	3.696936	28.203994
3	lapply+pmin	4.027896	3.923238	3.871584	3.928649	3.904770	2.936479
2	apply+min	6.969857	6.630471	7.040785	6.540463	6.659560	28.644606

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.251866	0.2734045	0.2879059	0.2838515	0.2923460	0.384646
4	lapply+pmin.int	0.581773	0.6260660	0.6624369	0.6448945	0.6725805	0.945455
3	lapply+pmin	0.638555	0.6904935	0.8486707	0.7131885	0.7489960	13.055857
2	apply+min	1.504383	1.6841305	1.9164179	1.7282255	1.8514770	14.720780

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	lapply+pmin.int	2.309851	2.289889	2.300880	2.271943	2.300632	2.457987
3	lapply+pmin	2.535296	2.525538	2.947737	2.512541	2.562019	33.942526
2	apply+min	5.972950	6.159849	6.656404	6.088485	6.333170	38.270982

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on double+100x1000 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	227.381	254.8190	273.9133	267.3340	277.895	502.656
2	rowMins	251.866	273.4045	287.9059	283.8515	292.346	384.646

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowMins	1.107683	1.072936	1.051084	1.061786	1.052002	0.7652271

Figure: Benchmarking of colMins() and rowMins() 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 5252191 280.5    8529671 455.6  8529671 455.6
Vcells 9970895  76.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMins = colMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 2L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(nrow(X)), function(i) X[i, 
+     ])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(nrow(X)), function(i) X[i, ])), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5252185 280.5    8529671 455.6  8529671 455.6
Vcells 10071002  76.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMins = rowMins(X, na.rm = FALSE), `apply+min` = apply(X, MARGIN = 1L, 
+     FUN = min, na.rm = FALSE), `lapply+pmin` = do.call(pmin, lapply(seq_len(ncol(X)), function(i) X[, 
+     i])), `lapply+pmin.int` = do.call(pmin.int, lapply(seq_len(ncol(X)), function(i) X[, i])), unit = "ms")

Table: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	colMins	0.198985	0.217623	0.2302715	0.226118	0.235797	0.318356
2	apply+min	0.784831	0.852859	1.0999258	0.877416	0.908524	19.860580
4	lapply+pmin.int	1.791695	1.976490	2.0853159	2.025352	2.102216	3.437286
3	lapply+pmin	2.259539	2.573584	2.8327318	2.656083	2.724118	19.084852

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
2	apply+min	3.944172	3.918975	4.776647	3.880346	3.852992	62.38481
4	lapply+pmin.int	9.004171	9.082176	9.055900	8.957056	8.915364	10.79699
3	lapply+pmin	11.355323	11.825882	12.301702	11.746446	11.552810	59.94815

Table: Benchmarking of rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() 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	rowMins	0.230426	0.2557065	0.2684111	0.2635605	0.2765010	0.361881
2	apply+min	0.803615	0.8602285	1.0989035	0.8957240	0.9356645	18.147457
4	lapply+pmin.int	1.458839	1.6289240	1.7163720	1.6748620	1.7383040	2.867851
3	lapply+pmin	1.940831	2.1925340	2.4580825	2.2610000	2.3534130	19.545645

	expr	min	lq	mean	median	uq	max
1	rowMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	apply+min	3.487519	3.364124	4.094106	3.398552	3.383946	50.147582
4	lapply+pmin.int	6.331052	6.370288	6.394564	6.354754	6.286791	7.924845
3	lapply+pmin	8.422795	8.574416	9.157902	8.578675	8.511409	54.011250

Figure: Benchmarking of colMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on double+1000x100 data as well as rowMins(), apply+min(), lapply+pmin() and lapply+pmin.int() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colMins() and rowMins() 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	colMins	198.985	217.6230	230.2715	226.1180	235.797	318.356
2	rowMins	230.426	255.7065	268.4111	263.5605	276.501	361.881

	expr	min	lq	mean	median	uq	max
1	colMins	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins	1.158007	1.174998	1.165629	1.165588	1.172623	1.136718

Figure: Benchmarking of colMins() and rowMins() 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 29.15 secs.

Reproducibility

To reproduce this report, do:

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

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 19:01:20 (+0200 UTC). Powered by RSP.