matrixStats.benchmarks

colMedians() and rowMedians() benchmarks on subsetted computation

This report benchmark the performance of colMedians() and rowMedians() on subsetted computation.

Data type “integer”

Data

> rmatrix <- function(nrow, ncol, mode = c("logical", "double", "integer", "index"), range = c(-100, 
+     +100), na_prob = 0) {
+     mode <- match.arg(mode)
+     n <- nrow * ncol
+     if (mode == "logical") {
+         x <- sample(c(FALSE, TRUE), size = n, replace = TRUE)
+     }     else if (mode == "index") {
+         x <- seq_len(n)
+         mode <- "integer"
+     }     else {
+         x <- runif(n, min = range[1], max = range[2])
+     }
+     storage.mode(x) <- mode
+     if (na_prob > 0) 
+         x[sample(n, size = na_prob * n)] <- NA
+     dim(x) <- c(nrow, ncol)
+     x
+ }
> rmatrices <- function(scale = 10, seed = 1, ...) {
+     set.seed(seed)
+     data <- list()
+     data[[1]] <- rmatrix(nrow = scale * 1, ncol = scale * 1, ...)
+     data[[2]] <- rmatrix(nrow = scale * 10, ncol = scale * 10, ...)
+     data[[3]] <- rmatrix(nrow = scale * 100, ncol = scale * 1, ...)
+     data[[4]] <- t(data[[3]])
+     data[[5]] <- rmatrix(nrow = scale * 10, ncol = scale * 100, ...)
+     data[[6]] <- t(data[[5]])
+     names(data) <- sapply(data, FUN = function(x) paste(dim(x), collapse = "x"))
+     data
+ }
> data <- rmatrices(mode = mode)

Results

10x10 integer matrix

> X <- data[["10x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5261358 281.0    7916910 422.9  7916910 422.9
Vcells 10162638  77.6   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5248479 280.3    7916910 422.9  7916910 422.9
Vcells 10120101  77.3   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.002539	0.0026150	0.0029159	0.0026825	0.0028075	0.020596
2	colMedians(X, rows, cols)	0.002845	0.0029630	0.0030915	0.0030270	0.0031540	0.005979
3	colMedians(X[rows, cols])	0.003396	0.0036455	0.0038512	0.0037765	0.0038855	0.008260

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colMedians(X, rows, cols)	1.120520	1.133078	1.060207	1.128425	1.123419	0.2902991
3	colMedians(X[rows, cols])	1.337535	1.394073	1.320767	1.407828	1.383971	0.4010487

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.002582	0.0026555	0.0027794	0.0027055	0.0028280	0.005478
2	rowMedians(X, cols, rows)	0.002927	0.0030785	0.0033691	0.0031330	0.0032205	0.022349
3	rowMedians(X[cols, rows])	0.003506	0.0037010	0.0038899	0.0037835	0.0039360	0.006272

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians(X, cols, rows)	1.133617	1.159292	1.212178	1.158011	1.138791	4.079774
3	rowMedians(X[cols, rows])	1.357862	1.393711	1.399545	1.398448	1.391796	1.144943

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

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	2.539	2.6150	2.91592	2.6825	2.8075	20.596
2	rowMedians_X_S	2.582	2.6555	2.77941	2.7055	2.8280	5.478

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.000000
2	rowMedians_X_S	1.016936	1.015488	0.9531846	1.008574	1.007302	0.265974

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

100x100 integer matrix

> X <- data[["100x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5247066 280.3    7916910 422.9  7916910 422.9
Vcells 9788306  74.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5247060 280.3    7916910 422.9  7916910 422.9
Vcells 9793389  74.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.067030	0.0697475	0.0793841	0.0764250	0.0835475	0.137287
2	colMedians(X, rows, cols)	0.068161	0.0714950	0.0804575	0.0780000	0.0861465	0.133461
3	colMedians(X[rows, cols])	0.077466	0.0805995	0.0896371	0.0872435	0.0966095	0.148208

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colMedians(X, rows, cols)	1.016873	1.025055	1.013521	1.020608	1.031108	0.9721314
3	colMedians(X[rows, cols])	1.155692	1.155590	1.129157	1.141557	1.156342	1.0795487

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.064076	0.068283	0.0754529	0.0738020	0.0816530	0.105578
2	rowMedians(X, cols, rows)	0.067314	0.070207	0.0774424	0.0762515	0.0825315	0.125626
3	rowMedians(X[cols, rows])	0.074273	0.078959	0.0861093	0.0847715	0.0915495	0.109607

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians(X, cols, rows)	1.050534	1.028177	1.026368	1.033190	1.010759	1.189888
3	rowMedians(X[cols, rows])	1.159139	1.156349	1.141233	1.148634	1.121202	1.038161

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

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

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	64.076	68.2830	75.45288	73.802	81.6530	105.578
1	colMedians_X_S	67.030	69.7475	79.38413	76.425	83.5475	137.287

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colMedians_X_S	1.046102	1.021448	1.052102	1.035541	1.023202	1.300337

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

1000x10 integer matrix

> X <- data[["1000x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5247808 280.3    7916910 422.9  7916910 422.9
Vcells 9792357  74.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5247802 280.3    7916910 422.9  7916910 422.9
Vcells 9797440  74.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.051732	0.0562565	0.0601103	0.0592305	0.0639270	0.076885
2	colMedians(X, rows, cols)	0.053468	0.0567755	0.0621583	0.0610460	0.0652130	0.085031
3	colMedians(X[rows, cols])	0.061557	0.0647875	0.0720022	0.0692910	0.0763975	0.125133

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMedians(X, rows, cols)	1.033558	1.009226	1.034070	1.030651	1.020117	1.105950
3	colMedians(X[rows, cols])	1.189921	1.151645	1.197834	1.169853	1.195074	1.627535

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.051167	0.055490	0.0597463	0.0583290	0.0625435	0.083070
2	rowMedians(X, cols, rows)	0.053707	0.058568	0.0627956	0.0630935	0.0667320	0.076504
3	rowMedians(X[cols, rows])	0.063305	0.068806	0.0744493	0.0742565	0.0788030	0.131429

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowMedians(X, cols, rows)	1.049641	1.055470	1.051037	1.081683	1.066969	0.9209582
3	rowMedians(X[cols, rows])	1.237223	1.239971	1.246091	1.273063	1.259971	1.5821476

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

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

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	51.167	55.4900	59.74629	58.3290	62.5435	83.070
1	colMedians_X_S	51.732	56.2565	60.11033	59.2305	63.9270	76.885

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colMedians_X_S	1.011042	1.013813	1.006093	1.015455	1.022121	0.9255447

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

10x1000 integer matrix

> X <- data[["10x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248012 280.3    7916910 422.9  7916910 422.9
Vcells 9793238  74.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248006 280.3    7916910 422.9  7916910 422.9
Vcells 9798321  74.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.053946	0.0605370	0.0657424	0.065430	0.069156	0.110516
2	colMedians(X, rows, cols)	0.060174	0.0642105	0.0707670	0.069485	0.075568	0.101287
3	colMedians(X[rows, cols])	0.065230	0.0734625	0.0791696	0.079116	0.084034	0.099485

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colMedians(X, rows, cols)	1.115449	1.060682	1.076428	1.061975	1.092718	0.9164917
3	colMedians(X[rows, cols])	1.209172	1.213514	1.204239	1.209170	1.215137	0.9001864

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.055236	0.0613350	0.0655234	0.0646645	0.0693605	0.092072
2	rowMedians(X, cols, rows)	0.056363	0.0670275	0.0725988	0.0713910	0.0785210	0.118767
3	rowMedians(X[cols, rows])	0.064038	0.0714920	0.0778549	0.0770710	0.0842430	0.099946

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians(X, cols, rows)	1.020403	1.092810	1.107983	1.104021	1.132071	1.289936
3	rowMedians(X[cols, rows])	1.159353	1.165599	1.188201	1.191860	1.214567	1.085520

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

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

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	55.236	61.335	65.52338	64.6645	69.3605	92.072
1	colMedians_X_S	53.946	60.537	65.74241	65.4300	69.1560	110.516

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	1.0000000	1.0000000	1.000000	1.000000	1.0000000	1.000000
1	colMedians_X_S	0.9766457	0.9869895	1.003343	1.011838	0.9970516	1.200322

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

100x1000 integer matrix

> X <- data[["100x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248225 280.3    7916910 422.9  7916910 422.9
Vcells 9815924  74.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248219 280.3    7916910 422.9  7916910 422.9
Vcells 9866007  75.3   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.617813	0.6192955	0.6433942	0.6204390	0.625912	0.866421
2	colMedians(X, rows, cols)	0.625488	0.6277925	0.6480555	0.6292565	0.635836	0.913809
3	colMedians(X[rows, cols])	0.685097	0.6870250	0.7164150	0.6883195	0.703381	0.916572

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMedians(X, rows, cols)	1.012423	1.013720	1.007245	1.014212	1.015855	1.054694
3	colMedians(X[rows, cols])	1.108907	1.109365	1.113493	1.109407	1.123770	1.057883

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.617588	0.6217030	0.6845461	0.6335880	0.6729925	1.095878
2	rowMedians(X, cols, rows)	0.639315	0.6420035	0.6794513	0.6485915	0.6680865	1.266033
3	rowMedians(X[cols, rows])	0.682908	0.6888280	0.7544767	0.7046215	0.7500345	1.149237

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.0000000	1.000000	1.0000000	1.000000
2	rowMedians(X, cols, rows)	1.035180	1.032653	0.9925574	1.023680	0.9927102	1.155268
3	rowMedians(X[cols, rows])	1.105766	1.107970	1.1021562	1.112113	1.1144768	1.048691

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

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	617.813	619.2955	643.3942	620.439	625.9120	866.421
2	rowMedians_X_S	617.588	621.7030	684.5461	633.588	672.9925	1095.878

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.0000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians_X_S	0.9996358	1.003888	1.063961	1.021193	1.075219	1.264833

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

1000x100 integer matrix

> X <- data[["1000x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248435 280.3    7916910 422.9  7916910 422.9
Vcells 9816719  74.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248429 280.3    7916910 422.9  7916910 422.9
Vcells 9866802  75.3   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.524673	0.5284170	0.6080464	0.5625435	0.6194565	1.118931
2	colMedians(X, rows, cols)	0.531137	0.5425495	0.6333145	0.5776520	0.6754200	1.172393
3	colMedians(X[rows, cols])	0.588329	0.6008250	0.7059546	0.6505860	0.7684900	1.324848

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
2	colMedians(X, rows, cols)	1.012320	1.026745	1.041556	1.026857	1.090343	1.04778
3	colMedians(X[rows, cols])	1.121325	1.137028	1.161021	1.156508	1.240587	1.18403

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.526715	0.5290245	0.5710529	0.5335425	0.5847075	0.836033
2	rowMedians(X, cols, rows)	0.551788	0.5546635	0.5970884	0.5592390	0.5949400	0.969852
3	rowMedians(X[cols, rows])	0.595780	0.6022510	0.6614477	0.6136995	0.6897845	1.015102

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians(X, cols, rows)	1.047603	1.048465	1.045592	1.048162	1.017500	1.160064
3	rowMedians(X[cols, rows])	1.131124	1.138418	1.158295	1.150235	1.179709	1.214189

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

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

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	526.715	529.0245	571.0529	533.5425	584.7075	836.033
1	colMedians_X_S	524.673	528.4170	608.0465	562.5435	619.4565	1118.931

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	1.0000000	1.0000000	1.000000	1.000000	1.00000	1.000000
1	colMedians_X_S	0.9961231	0.9988517	1.064781	1.054356	1.05943	1.338381

Figure: Benchmarking of colMedians_X_S() and rowMedians_X_S() on integer+1000x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

Data type “double”

Data

> rmatrix <- function(nrow, ncol, mode = c("logical", "double", "integer", "index"), range = c(-100, 
+     +100), na_prob = 0) {
+     mode <- match.arg(mode)
+     n <- nrow * ncol
+     if (mode == "logical") {
+         x <- sample(c(FALSE, TRUE), size = n, replace = TRUE)
+     }     else if (mode == "index") {
+         x <- seq_len(n)
+         mode <- "integer"
+     }     else {
+         x <- runif(n, min = range[1], max = range[2])
+     }
+     storage.mode(x) <- mode
+     if (na_prob > 0) 
+         x[sample(n, size = na_prob * n)] <- NA
+     dim(x) <- c(nrow, ncol)
+     x
+ }
> rmatrices <- function(scale = 10, seed = 1, ...) {
+     set.seed(seed)
+     data <- list()
+     data[[1]] <- rmatrix(nrow = scale * 1, ncol = scale * 1, ...)
+     data[[2]] <- rmatrix(nrow = scale * 10, ncol = scale * 10, ...)
+     data[[3]] <- rmatrix(nrow = scale * 100, ncol = scale * 1, ...)
+     data[[4]] <- t(data[[3]])
+     data[[5]] <- rmatrix(nrow = scale * 10, ncol = scale * 100, ...)
+     data[[6]] <- t(data[[5]])
+     names(data) <- sapply(data, FUN = function(x) paste(dim(x), collapse = "x"))
+     data
+ }
> data <- rmatrices(mode = mode)

Results

10x10 double matrix

> X <- data[["10x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248654 280.4    7916910 422.9  7916910 422.9
Vcells 9907851  75.6   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248639 280.4    7916910 422.9  7916910 422.9
Vcells 9908019  75.6   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.002866	0.0029550	0.0032791	0.0030360	0.0031385	0.022626
2	colMedians(X, rows, cols)	0.003050	0.0032545	0.0034643	0.0033360	0.0034415	0.010705
3	colMedians(X[rows, cols])	0.003830	0.0040550	0.0043854	0.0041545	0.0042640	0.019014

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colMedians(X, rows, cols)	1.064201	1.101354	1.056488	1.098814	1.096543	0.4731283
3	colMedians(X[rows, cols])	1.336357	1.372250	1.337369	1.368412	1.358611	0.8403606

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.002877	0.0030065	0.0031335	0.0030585	0.0031765	0.006756
2	rowMedians(X, cols, rows)	0.003057	0.0033045	0.0036690	0.0033760	0.0035090	0.027124
3	rowMedians(X[cols, rows])	0.003871	0.0041005	0.0042870	0.0041935	0.0043220	0.007832

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians(X, cols, rows)	1.062565	1.099119	1.170893	1.103809	1.104675	4.014802
3	rowMedians(X[cols, rows])	1.345499	1.363878	1.368101	1.371097	1.360617	1.159266

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

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	2.866	2.9550	3.27911	3.0360	3.1385	22.626
2	rowMedians_X_S	2.877	3.0065	3.13354	3.0585	3.1765	6.756

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	rowMedians_X_S	1.003838	1.017428	0.9556069	1.007411	1.012108	0.2985945

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

100x100 double matrix

> X <- data[["100x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248852 280.4    7916910 422.9  7916910 422.9
Vcells 9913818  75.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5248846 280.4    7916910 422.9  7916910 422.9
Vcells 9923901  75.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.103994	0.110689	0.1200387	0.1141755	0.1265470	0.157468
2	colMedians(X, rows, cols)	0.106586	0.111645	0.1205535	0.1155800	0.1283115	0.152400
3	colMedians(X[rows, cols])	0.116216	0.123479	0.1347654	0.1280850	0.1460190	0.195296

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colMedians(X, rows, cols)	1.024925	1.008637	1.004289	1.012301	1.013943	0.9678157
3	colMedians(X[rows, cols])	1.117526	1.115549	1.122683	1.121826	1.153872	1.2402266

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.109658	0.1139415	0.1283539	0.1256335	0.1385855	0.179993
2	rowMedians(X, cols, rows)	0.110766	0.1156400	0.1304482	0.1278560	0.1392520	0.201395
3	rowMedians(X[cols, rows])	0.122515	0.1281795	0.1444028	0.1414955	0.1576605	0.193491

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians(X, cols, rows)	1.010104	1.014907	1.016317	1.017690	1.004809	1.118905
3	rowMedians(X[cols, rows])	1.117246	1.124959	1.125037	1.126256	1.137641	1.074992

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

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	103.994	110.6890	120.0387	114.1755	126.5470	157.468
2	rowMedians_X_S	109.658	113.9415	128.3539	125.6335	138.5855	179.993

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians_X_S	1.054465	1.029384	1.069271	1.100354	1.095131	1.143045

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

1000x10 double matrix

> X <- data[["1000x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5249051 280.4    7916910 422.9  7916910 422.9
Vcells 9915238  75.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5249045 280.4    7916910 422.9  7916910 422.9
Vcells 9925321  75.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
2	colMedians(X, rows, cols)	0.101813	0.108786	0.1170997	0.1125975	0.1253585	0.152606
1	colMedians_X_S	0.099749	0.107163	0.1173915	0.1126875	0.1271455	0.141702
3	colMedians(X[rows, cols])	0.112376	0.116818	0.1292413	0.1242275	0.1409765	0.192323

	expr	min	lq	mean	median	uq	max
2	colMedians(X, rows, cols)	1.0000000	1.0000000	1.000000	1.000000	1.000000	1.000000
1	colMedians_X_S	0.9797275	0.9850808	1.002492	1.000799	1.014255	0.928548
3	colMedians(X[rows, cols])	1.1037490	1.0738330	1.103687	1.103288	1.124587	1.260258

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.097308	0.1070735	0.1136451	0.1103630	0.1178175	0.144058
2	rowMedians(X, cols, rows)	0.100292	0.1094890	0.1177984	0.1139905	0.1283830	0.155317
3	rowMedians(X[cols, rows])	0.112283	0.1227505	0.1311003	0.1270895	0.1399430	0.200809

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians(X, cols, rows)	1.030666	1.022559	1.036546	1.032869	1.089677	1.078156
3	rowMedians(X[cols, rows])	1.153893	1.146413	1.153593	1.151559	1.187795	1.393946

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

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

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	97.308	107.0735	113.6451	110.3630	117.8175	144.058
1	colMedians_X_S	99.749	107.1630	117.3915	112.6875	127.1455	141.702

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colMedians_X_S	1.025085	1.000836	1.032965	1.021062	1.079173	0.9836455

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

10x1000 double matrix

> X <- data[["10x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5249255 280.4    7916910 422.9  7916910 422.9
Vcells 9915373  75.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5249249 280.4    7916910 422.9  7916910 422.9
Vcells 9925456  75.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.083126	0.0901515	0.0967791	0.0940785	0.1018675	0.147479
2	colMedians(X, rows, cols)	0.085603	0.0920450	0.0980304	0.0949215	0.1020990	0.137271
3	colMedians(X[rows, cols])	0.101828	0.1068940	0.1134429	0.1098645	0.1177940	0.156936

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colMedians(X, rows, cols)	1.029798	1.021003	1.012930	1.008961	1.002273	0.9307834
3	colMedians(X[rows, cols])	1.224984	1.185715	1.172184	1.167796	1.156345	1.0641244

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.083556	0.0913700	0.0999967	0.0953660	0.1068230	0.167358
2	rowMedians(X, cols, rows)	0.086561	0.0925225	0.0998162	0.0959085	0.1024865	0.161939
3	rowMedians(X[cols, rows])	0.096264	0.1044415	0.1163377	0.1112965	0.1236995	0.204862

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.0000000	1.000000	1.0000000	1.0000000
2	rowMedians(X, cols, rows)	1.035964	1.012614	0.9981943	1.005689	0.9594048	0.9676203
3	rowMedians(X[cols, rows])	1.152090	1.143061	1.1634150	1.167046	1.1579856	1.2240945

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

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	83.126	90.1515	96.77906	94.0785	101.8675	147.479
2	rowMedians_X_S	83.556	91.3700	99.99674	95.3660	106.8230	167.358

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians_X_S	1.005173	1.013516	1.033248	1.013685	1.048647	1.134792

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

100x1000 double matrix

> X <- data[["100x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5249468 280.4    7916910 422.9  7916910 422.9
Vcells 9960869  76.0   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5249462 280.4    7916910 422.9  7916910 422.9
Vcells 10060952  76.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.899679	0.9511215	1.038798	0.9962655	1.074001	1.561260
2	colMedians(X, rows, cols)	0.913745	0.9796190	1.079340	1.0483620	1.136159	1.808542
3	colMedians(X[rows, cols])	0.996477	1.0448190	1.199718	1.1293160	1.288202	1.919176

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMedians(X, rows, cols)	1.015634	1.029962	1.039027	1.052292	1.057875	1.158386
3	colMedians(X[rows, cols])	1.107592	1.098513	1.154909	1.133549	1.199442	1.229248

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.917956	1.004488	1.117562	1.072184	1.185041	1.739220
2	rowMedians(X, cols, rows)	0.940122	1.036550	1.175913	1.130084	1.273694	2.034522
3	rowMedians(X[cols, rows])	1.002727	1.095776	1.236680	1.174289	1.328539	1.753591

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians(X, cols, rows)	1.024147	1.031919	1.052212	1.054002	1.074810	1.169790
3	rowMedians(X[cols, rows])	1.092348	1.090880	1.106587	1.095231	1.121091	1.008263

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

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	899.679	951.1215	1038.799	996.2655	1074.001	1561.26
2	rowMedians_X_S	917.956	1004.4880	1117.562	1072.1840	1185.041	1739.22

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians_X_S	1.020315	1.056109	1.075822	1.076203	1.103389	1.113985

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

1000x100 double matrix

> X <- data[["1000x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5249678 280.4    7916910 422.9  7916910 422.9
Vcells 9961010  76.0   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMedians_X_S = colMedians(X_S, na.rm = FALSE), `colMedians(X, rows, cols)` = colMedians(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMedians(X[rows, cols])` = colMedians(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5249672 280.4    7916910 422.9  7916910 422.9
Vcells 10061093  76.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMedians_X_S = rowMedians(X_S, na.rm = FALSE), `rowMedians(X, cols, rows)` = rowMedians(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMedians(X[cols, rows])` = rowMedians(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	0.826431	0.875139	0.9632611	0.9171335	1.010642	1.482132
2	colMedians(X, rows, cols)	0.834494	0.927260	1.0432954	0.9840365	1.090420	1.895048
3	colMedians(X[rows, cols])	0.919544	1.021737	1.1578928	1.1212815	1.274457	2.079346

	expr	min	lq	mean	median	uq	max
1	colMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMedians(X, rows, cols)	1.009756	1.059557	1.083087	1.072948	1.078937	1.278596
3	colMedians(X[rows, cols])	1.112669	1.167515	1.202055	1.222594	1.261037	1.402942

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

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	0.827908	0.8323610	0.8929356	0.8347095	0.8697840	1.533365
2	rowMedians(X, cols, rows)	0.864646	0.8697185	0.9246822	0.8722690	0.8823855	1.538825
3	rowMedians(X[cols, rows])	0.928314	0.9371655	1.0057969	0.9427665	0.9748030	1.642082

	expr	min	lq	mean	median	uq	max
1	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMedians(X, cols, rows)	1.044374	1.044881	1.035553	1.044997	1.014488	1.003561
3	rowMedians(X[cols, rows])	1.121277	1.125912	1.126394	1.129455	1.120742	1.070901

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

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

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	827.908	832.361	892.9356	834.7095	869.784	1533.365
1	colMedians_X_S	826.431	875.139	963.2611	917.1335	1010.642	1482.132

	expr	min	lq	mean	median	uq	max
2	rowMedians_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colMedians_X_S	0.998216	1.051394	1.078758	1.098746	1.161946	0.9665879

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

Appendix

Session information

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

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

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

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

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

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

Total processing time was 25.56 secs.

Reproducibility

To reproduce this report, do:

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