matrixStats.benchmarks

---

colMins() and rowMins() benchmarks on subsetted computation

This report benchmark the performance of colMins() and rowMins() 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  5268973 281.4    7916910 422.9  7916910 422.9
Vcells 10216448  78.0   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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  5256234 280.8    7916910 422.9  7916910 422.9
Vcells 10174214  77.7   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.002035	0.0021205	0.0043266	0.0021885	0.0024690	0.179827
2	colMins(X, rows, cols)	0.002415	0.0025015	0.0029077	0.0026065	0.0027655	0.005682
3	colMins(X[rows, cols])	0.002798	0.0031070	0.0037563	0.0032395	0.0035045	0.010028

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	colMins(X, rows, cols)	1.186732	1.179675	0.6720504	1.190998	1.120089	0.0315970
3	colMins(X[rows, cols])	1.374939	1.465221	0.8681892	1.480238	1.419401	0.0557647

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.003193	0.0034715	0.0036700	0.0036100	0.0037365	0.006445
2	rowMins(X, cols, rows)	0.003682	0.0039600	0.0065779	0.0041345	0.0043695	0.226264
3	rowMins(X[cols, rows])	0.004535	0.0050790	0.0055550	0.0053325	0.0057570	0.012191

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.153147	1.140717	1.792318	1.145291	1.169410	35.106905
3	rowMins(X[cols, rows])	1.420294	1.463056	1.513617	1.477147	1.540747	1.891544

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

Table: Benchmarking of colMins_X_S() and rowMins_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	colMins_X_S	2.035	2.1205	4.32658	2.1885	2.4690	179.827
2	rowMins_X_S	3.193	3.4715	3.67003	3.6100	3.7365	6.445

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
2	rowMins_X_S	1.569042	1.637114	0.848252	1.649532	1.513366	0.03584

Figure: Benchmarking of colMins_X_S() and rowMins_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 5254937 280.7    7916910 422.9  7916910 422.9
Vcells 9843298  75.1   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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 5254931 280.7    7916910 422.9  7916910 422.9
Vcells 9848381  75.2   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.013694	0.0141990	0.0153610	0.0145165	0.0149595	0.031697
2	colMins(X, rows, cols)	0.017540	0.0180960	0.0201012	0.0187155	0.0195865	0.049373
3	colMins(X[rows, cols])	0.023019	0.0238335	0.0261700	0.0245290	0.0257800	0.070284

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMins(X, rows, cols)	1.280853	1.274456	1.308585	1.289257	1.309302	1.557655
3	colMins(X[rows, cols])	1.680955	1.678534	1.703663	1.689732	1.723320	2.217371

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.016432	0.017082	0.0177373	0.0173750	0.0179165	0.027156
2	rowMins(X, cols, rows)	0.019507	0.020242	0.0218397	0.0206080	0.0212705	0.059040
3	rowMins(X[cols, rows])	0.027337	0.027821	0.0289056	0.0283285	0.0292650	0.042700

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.187135	1.184990	1.231288	1.186072	1.187202	2.174105
3	rowMins(X[cols, rows])	1.663644	1.628674	1.629650	1.630417	1.633411	1.572396

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

Table: Benchmarking of colMins_X_S() and rowMins_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
1	colMins_X_S	13.694	14.199	15.36103	14.5165	14.9595	31.697
2	rowMins_X_S	16.432	17.082	17.73731	17.3750	17.9165	27.156

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowMins_X_S	1.199942	1.203043	1.154695	1.196914	1.197667	0.8567372

Figure: Benchmarking of colMins_X_S() and rowMins_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 5255679 280.7    7916910 422.9  7916910 422.9
Vcells 9847340  75.2   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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 5255673 280.7    7916910 422.9  7916910 422.9
Vcells 9852423  75.2   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.012131	0.0241615	0.0232703	0.0247145	0.0252980	0.039351
2	colMins(X, rows, cols)	0.015570	0.0311600	0.0302342	0.0316410	0.0325550	0.049335
3	colMins(X[rows, cols])	0.022870	0.0448585	0.0434025	0.0453020	0.0471425	0.081244

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMins(X, rows, cols)	1.283489	1.289655	1.299261	1.280261	1.286861	1.253717
3	colMins(X[rows, cols])	1.885253	1.856611	1.865144	1.833013	1.863487	2.064598

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.012845	0.0131460	0.0139520	0.0133030	0.0134370	0.027972
2	rowMins(X, cols, rows)	0.016323	0.0165285	0.0169706	0.0166870	0.0169000	0.026606
3	rowMins(X[cols, rows])	0.024136	0.0244920	0.0254436	0.0246795	0.0249525	0.071586

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowMins(X, cols, rows)	1.270767	1.257303	1.216359	1.254379	1.257721	0.9511655
3	rowMins(X[cols, rows])	1.879019	1.863076	1.823656	1.855183	1.856999	2.5592021

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

Table: Benchmarking of colMins_X_S() and rowMins_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	rowMins_X_S	12.845	13.1460	13.95198	13.3030	13.437	27.972
1	colMins_X_S	12.131	24.1615	23.27031	24.7145	25.298	39.351

	expr	min	lq	mean	median	uq	max
2	rowMins_X_S	1.0000000	1.000000	1.000000	1.000000	1.000000	1.0000
1	colMins_X_S	0.9444142	1.837935	1.667886	1.857814	1.882712	1.4068

Figure: Benchmarking of colMins_X_S() and rowMins_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 5255883 280.7    7916910 422.9  7916910 422.9
Vcells 9848159  75.2   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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 5255877 280.7    7916910 422.9  7916910 422.9
Vcells 9853242  75.2   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.021281	0.0220745	0.0248124	0.0227165	0.0242525	0.060299
2	colMins(X, rows, cols)	0.024593	0.0252180	0.0283586	0.0259455	0.0268240	0.067430
3	colMins(X[rows, cols])	0.030502	0.0316420	0.0362147	0.0326805	0.0368075	0.059531

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colMins(X, rows, cols)	1.155632	1.142404	1.142919	1.142143	1.106030	1.1182607
3	colMins(X[rows, cols])	1.433297	1.433419	1.459539	1.438624	1.517679	0.9872635

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.021779	0.0228345	0.0243946	0.0235275	0.0243565	0.060428
2	rowMins(X, cols, rows)	0.027299	0.0288975	0.0306052	0.0296830	0.0301040	0.063419
3	rowMins(X[cols, rows])	0.031499	0.0326040	0.0336025	0.0332335	0.0338295	0.048796

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowMins(X, cols, rows)	1.253455	1.265519	1.254586	1.261630	1.235974	1.0494969
3	rowMins(X[cols, rows])	1.446301	1.427839	1.377456	1.412538	1.388931	0.8075065

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

Table: Benchmarking of colMins_X_S() and rowMins_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
1	colMins_X_S	21.281	22.0745	24.81245	22.7165	24.2525	60.299
2	rowMins_X_S	21.779	22.8345	24.39463	23.5275	24.3565	60.428

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.000000
2	rowMins_X_S	1.023401	1.034429	0.9831609	1.035701	1.004288	1.002139

Figure: Benchmarking of colMins_X_S() and rowMins_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 5256096 280.8    7916910 422.9  7916910 422.9
Vcells 9870825  75.4   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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 5256090 280.8    7916910 422.9  7916910 422.9
Vcells 9920908  75.7   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.110041	0.118681	0.1325330	0.1268325	0.1416945	0.251385
2	colMins(X, rows, cols)	0.137480	0.147516	0.1631676	0.1559895	0.1750660	0.265372
3	colMins(X[rows, cols])	0.188574	0.205857	0.2344054	0.2208370	0.2430135	0.509236

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMins(X, rows, cols)	1.249353	1.242962	1.231146	1.229886	1.235517	1.055640
3	colMins(X[rows, cols])	1.713670	1.734541	1.768657	1.741170	1.715053	2.025721

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.115127	0.1255520	0.1358057	0.1305705	0.1400010	0.239547
2	rowMins(X, cols, rows)	0.135884	0.1446415	0.1612695	0.1535720	0.1663995	0.278447
3	rowMins(X[cols, rows])	0.186785	0.2042905	0.2251474	0.2156145	0.2345030	0.417285

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.180297	1.152045	1.187502	1.176162	1.188559	1.162390
3	rowMins(X[cols, rows])	1.622426	1.627139	1.657865	1.651326	1.675010	1.741975

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

Table: Benchmarking of colMins_X_S() and rowMins_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	colMins_X_S	110.041	118.681	132.5330	126.8325	141.6945	251.385
2	rowMins_X_S	115.127	125.552	135.8057	130.5705	140.0010	239.547

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.0000000	1.0000000
2	rowMins_X_S	1.046219	1.057895	1.024693	1.029472	0.9880482	0.9529089

Figure: Benchmarking of colMins_X_S() and rowMins_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 5256306 280.8    7916910 422.9  7916910 422.9
Vcells 9871596  75.4   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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 5256300 280.8    7916910 422.9  7916910 422.9
Vcells 9921679  75.7   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.074310	0.0817675	0.1015839	0.0872670	0.101372	0.182153
2	colMins(X, rows, cols)	0.091755	0.1009640	0.1219759	0.1078535	0.121287	0.213577
3	colMins(X[rows, cols])	0.143374	0.1593630	0.2027460	0.1795150	0.246886	0.333111

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMins(X, rows, cols)	1.234760	1.234769	1.200740	1.235902	1.196455	1.172514
3	colMins(X[rows, cols])	1.929404	1.948977	1.995847	2.057078	2.435446	1.828743

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.094571	0.1014880	0.1111650	0.1088595	0.1165130	0.170661
2	rowMins(X, cols, rows)	0.115155	0.1253245	0.1340328	0.1307110	0.1358840	0.260444
3	rowMins(X[cols, rows])	0.169438	0.1845415	0.1986710	0.1942370	0.2105265	0.252262

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.217657	1.234870	1.205711	1.200731	1.166256	1.526090
3	rowMins(X[cols, rows])	1.791649	1.818358	1.787173	1.784291	1.806893	1.478147

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

Table: Benchmarking of colMins_X_S() and rowMins_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
1	colMins_X_S	74.310	81.7675	101.5839	87.2670	101.372	182.153
2	rowMins_X_S	94.571	101.4880	111.1650	108.8595	116.513	170.661

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.0000000
2	rowMins_X_S	1.272655	1.241178	1.094316	1.24743	1.149361	0.9369102

Figure: Benchmarking of colMins_X_S() and rowMins_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 5256524 280.8    7916910 422.9  7916910 422.9
Vcells 9962691  76.1   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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 5256509 280.8    7916910 422.9  7916910 422.9
Vcells 9962859  76.1   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.001837	0.0018845	0.0024692	0.0019720	0.0020620	0.030017
2	colMins(X, rows, cols)	0.002124	0.0021990	0.0024274	0.0022725	0.0024245	0.004709
3	colMins(X[rows, cols])	0.002675	0.0029210	0.0032631	0.0030290	0.0031390	0.012697

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	colMins(X, rows, cols)	1.156233	1.166888	0.9830715	1.152383	1.175800	0.1568778
3	colMins(X[rows, cols])	1.456179	1.550013	1.3215158	1.536004	1.522308	0.4229936

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.001881	0.0019400	0.0025046	0.0020475	0.0034925	0.004653
2	rowMins(X, cols, rows)	0.002113	0.0022285	0.0030957	0.0023540	0.0037055	0.032021
3	rowMins(X[cols, rows])	0.002647	0.0029060	0.0037384	0.0030655	0.0049565	0.008525

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.123339	1.148711	1.235991	1.149695	1.060988	6.881797
3	rowMins(X[cols, rows])	1.407230	1.497938	1.492604	1.497192	1.419184	1.832151

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

Table: Benchmarking of colMins_X_S() and rowMins_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	colMins_X_S	1.837	1.8845	2.46921	1.9720	2.0620	30.017
2	rowMins_X_S	1.881	1.9400	2.50463	2.0475	3.4925	4.653

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowMins_X_S	1.023952	1.029451	1.014345	1.038286	1.693744	0.1550122

Figure: Benchmarking of colMins_X_S() and rowMins_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 5256722 280.8    7916910 422.9  7916910 422.9
Vcells 9968629  76.1   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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 5256716 280.8    7916910 422.9  7916910 422.9
Vcells 9978712  76.2   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.008826	0.0096035	0.0104460	0.0099655	0.0104175	0.023403
2	colMins(X, rows, cols)	0.016654	0.0180305	0.0191589	0.0185955	0.0190825	0.032819
3	colMins(X[rows, cols])	0.023022	0.0240960	0.0262701	0.0246680	0.0254220	0.095363

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMins(X, rows, cols)	1.886925	1.877493	1.834093	1.865988	1.831773	1.402342
3	colMins(X[rows, cols])	2.608430	2.509085	2.514850	2.475340	2.440317	4.074820

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.016744	0.0177420	0.0183832	0.018066	0.0183355	0.028048
2	rowMins(X, cols, rows)	0.018139	0.0188500	0.0203472	0.019257	0.0197620	0.079672
3	rowMins(X[cols, rows])	0.031464	0.0327365	0.0342872	0.033136	0.0335485	0.075699

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.083313	1.062451	1.106836	1.065925	1.077800	2.840559
3	rowMins(X[cols, rows])	1.879121	1.845141	1.865135	1.834164	1.829702	2.698909

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

Table: Benchmarking of colMins_X_S() and rowMins_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	colMins_X_S	8.826	9.6035	10.44598	9.9655	10.4175	23.403
2	rowMins_X_S	16.744	17.7420	18.38324	18.0660	18.3355	28.048

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins_X_S	1.897122	1.847451	1.759839	1.812854	1.760067	1.198479

Figure: Benchmarking of colMins_X_S() and rowMins_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 5256921 280.8    7916910 422.9  7916910 422.9
Vcells 9970027  76.1   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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 5256915 280.8    7916910 422.9  7916910 422.9
Vcells 9980110  76.2   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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
1	colMins_X_S	0.007975	0.008264	0.0090750	0.0084315	0.0086075	0.022445
2	colMins(X, rows, cols)	0.014711	0.015122	0.0171978	0.0154695	0.0158285	0.030674
3	colMins(X[rows, cols])	0.022820	0.023247	0.0259668	0.0239035	0.0242010	0.071655

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMins(X, rows, cols)	1.844639	1.829865	1.895080	1.834727	1.838920	1.366629
3	colMins(X[rows, cols])	2.861442	2.813045	2.861362	2.835023	2.811618	3.192471

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.015009	0.0155105	0.0165006	0.0157930	0.0160465	0.027362
2	rowMins(X, cols, rows)	0.017131	0.0179790	0.0191335	0.0183005	0.0188200	0.033497
3	rowMins(X[cols, rows])	0.031724	0.0323215	0.0352720	0.0332405	0.0340575	0.085921

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.141382	1.159150	1.159566	1.158773	1.172841	1.224216
3	rowMins(X[cols, rows])	2.113665	2.083846	2.137622	2.104762	2.122425	3.140158

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

Table: Benchmarking of colMins_X_S() and rowMins_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
1	colMins_X_S	7.975	8.2640	9.07499	8.4315	8.6075	22.445
2	rowMins_X_S	15.009	15.5105	16.50056	15.7930	16.0465	27.362

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins_X_S	1.882006	1.876876	1.818245	1.873095	1.864246	1.219069

Figure: Benchmarking of colMins_X_S() and rowMins_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 5257125 280.8    7916910 422.9  7916910 422.9
Vcells 9970162  76.1   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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 5257119 280.8    7916910 422.9  7916910 422.9
Vcells 9980245  76.2   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.018037	0.0214340	0.0244914	0.023148	0.0263780	0.070856
2	colMins(X, rows, cols)	0.029407	0.0358425	0.0403879	0.037750	0.0420305	0.078423
3	colMins(X[rows, cols])	0.036084	0.0412110	0.0471068	0.043300	0.0511825	0.108447

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMins(X, rows, cols)	1.630371	1.672226	1.649063	1.630810	1.593392	1.106794
3	colMins(X[rows, cols])	2.000554	1.922693	1.923401	1.870572	1.940348	1.530527

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.025561	0.0272320	0.0311482	0.0285550	0.031266	0.049964
2	rowMins(X, cols, rows)	0.026252	0.0304295	0.0369572	0.0336225	0.038519	0.078634
3	rowMins(X[cols, rows])	0.040895	0.0432040	0.0485048	0.0451035	0.048643	0.073594

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.027033	1.117417	1.186496	1.177464	1.231977	1.573813
3	rowMins(X[cols, rows])	1.599898	1.586516	1.557228	1.579531	1.555779	1.472940

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

Table: Benchmarking of colMins_X_S() and rowMins_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	colMins_X_S	18.037	21.434	24.49141	23.148	26.378	70.856
2	rowMins_X_S	25.561	27.232	31.14820	28.555	31.266	49.964

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowMins_X_S	1.417142	1.270505	1.271801	1.233584	1.185306	0.7051485

Figure: Benchmarking of colMins_X_S() and rowMins_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  5257338 280.8    7916910 422.9  7916910 422.9
Vcells 10015606  76.5   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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  5257332 280.8    7916910 422.9  7916910 422.9
Vcells 10115689  77.2   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.086758	0.0959010	0.1153490	0.1084845	0.1264815	0.209635
2	colMins(X, rows, cols)	0.163238	0.1762275	0.2194041	0.2009445	0.2484930	0.464252
3	colMins(X[rows, cols])	0.202229	0.2249270	0.2879682	0.2726865	0.3361585	0.458204

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMins(X, rows, cols)	1.881533	1.837598	1.902090	1.852288	1.964659	2.214573
3	colMins(X[rows, cols])	2.330955	2.345408	2.496496	2.513599	2.657768	2.185723

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.122137	0.137787	0.1559285	0.1477625	0.1759455	0.234144
2	rowMins(X, cols, rows)	0.139865	0.152498	0.1723866	0.1673790	0.1845905	0.359318
3	rowMins(X[cols, rows])	0.217620	0.244753	0.2799769	0.2679955	0.3001220	0.430181

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.145148	1.106766	1.105549	1.132757	1.049135	1.534603
3	rowMins(X[cols, rows])	1.781770	1.776314	1.795547	1.813691	1.705767	1.837250

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

Table: Benchmarking of colMins_X_S() and rowMins_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	colMins_X_S	86.758	95.901	115.3490	108.4845	126.4815	209.635
2	rowMins_X_S	122.137	137.787	155.9285	147.7625	175.9455	234.144

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins_X_S	1.40779	1.436763	1.351798	1.362061	1.391077	1.116913

Figure: Benchmarking of colMins_X_S() and rowMins_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  5257548 280.8    7916910 422.9  7916910 422.9
Vcells 10015747  76.5   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colMins_X_S = colMins(X_S, na.rm = FALSE), `colMins(X, rows, cols)` = colMins(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colMins(X[rows, cols])` = colMins(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  5257542 280.8    7916910 422.9  7916910 422.9
Vcells 10115830  77.2   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowMins_X_S = rowMins(X_S, na.rm = FALSE), `rowMins(X, cols, rows)` = rowMins(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowMins(X[cols, rows])` = rowMins(X[cols, rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colMins_X_S(), colMins(X, rows, cols)() and colMins(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	colMins_X_S	0.052913	0.0565530	0.0608817	0.0586755	0.0624805	0.112666
2	colMins(X, rows, cols)	0.104066	0.1155635	0.1218716	0.1197615	0.1256415	0.165575
3	colMins(X[rows, cols])	0.160332	0.1713425	0.1807807	0.1785245	0.1848305	0.338876

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colMins(X, rows, cols)	1.966738	2.043455	2.001777	2.041082	2.010891	1.469609
3	colMins(X[rows, cols])	3.030106	3.029768	2.969378	3.042573	2.958211	3.007793

Table: Benchmarking of rowMins_X_S(), rowMins(X, cols, rows)() and rowMins(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	rowMins_X_S	0.123836	0.1284510	0.1471246	0.1328070	0.1432425	0.268057
2	rowMins(X, cols, rows)	0.134520	0.1441380	0.1781831	0.1504295	0.1958260	0.503150
3	rowMins(X[cols, rows])	0.243379	0.2537985	0.2974348	0.2649060	0.2928030	0.650017

	expr	min	lq	mean	median	uq	max
1	rowMins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins(X, cols, rows)	1.086275	1.122124	1.211103	1.132693	1.367094	1.877026
3	rowMins(X[cols, rows])	1.965333	1.975839	2.021652	1.994669	2.044107	2.424921

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

Table: Benchmarking of colMins_X_S() and rowMins_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
1	colMins_X_S	52.913	56.553	60.88169	58.6755	62.4805	112.666
2	rowMins_X_S	123.836	128.451	147.12465	132.8070	143.2425	268.057

	expr	min	lq	mean	median	uq	max
1	colMins_X_S	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowMins_X_S	2.34037	2.271338	2.416567	2.263415	2.292595	2.379218

Figure: Benchmarking of colMins_X_S() and rowMins_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 24.14 secs.

Reproducibility

To reproduce this report, do:

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

---

Copyright Dongcan Jiang. Last updated on 2021-08-25 22:21:22 (+0200 UTC). Powered by RSP.