colRowCounts_subset

matrixStats: Benchmark report

---

colCounts() and rowCounts() benchmarks on subsetted computation

This report benchmark the performance of colCounts() and rowCounts() on subsetted computation.

Data type "logical"

Data

> rmatrix <- function(nrow, ncol, mode = c("logical", "double", "integer", "index"), range = c(-100, 
+     +100), naProb = 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 (naProb > 0) 
+         X[sample(n, size = naProb * 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 matrix

> X <- data[["10x10"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

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

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	0.005384	0.0057790	0.0076935	0.0059465	0.0061550	0.170428	100
colCounts(X,rows,cols)	0.005801	0.0061375	0.0064140	0.0063020	0.0064855	0.008404	100
colCounts(X[rows,cols])	0.007176	0.0074875	0.0077561	0.0076095	0.0077755	0.014095	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000	100
colCounts(X,rows,cols)	1.077452	1.062035	0.8336938	1.059783	1.053696	0.0493111	100
colCounts(X[rows,cols])	1.332838	1.295639	1.0081302	1.279660	1.263282	0.0827035	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on logical+10x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	0.005357	0.0055655	0.0073684	0.0057185	0.0059035	0.163155	100
rowCounts(X,cols,rows)	0.005776	0.0059890	0.0062294	0.0061605	0.0063160	0.008153	100
rowCounts(X[cols,rows])	0.006979	0.0072510	0.0075992	0.0074155	0.0075820	0.012321	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000	100
rowCounts(X,cols,rows)	1.078215	1.076094	0.8454301	1.077293	1.069874	0.0499709	100
rowCounts(X[cols,rows])	1.302781	1.302848	1.0313312	1.296756	1.284323	0.0755171	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on logical+10x10 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+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	neval
2	rowCounts_X_S	5.357	5.5655	7.36838	5.7185	5.9035	163.155	100
1	colCounts_X_S	5.384	5.7790	7.69352	5.9465	6.1550	170.428	100

	expr	min	lq	mean	median	uq	max	neval
2	rowCounts_X_S	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000	100
1	colCounts_X_S	1.00504	1.038361	1.044126	1.039871	1.042602	1.044577	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+10x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x100 matrix

> X <- data[["100x100"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

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

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	0.018947	0.019930	0.0208104	0.0205495	0.0212630	0.029576	100
colCounts(X,rows,cols)	0.029228	0.030619	0.0312689	0.0309315	0.0311545	0.063351	100
colCounts(X[rows,cols])	0.035158	0.036291	0.0372238	0.0369725	0.0375915	0.051626	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.542619	1.536327	1.502561	1.505219	1.465198	2.141973	100
colCounts(X[rows,cols])	1.855597	1.820923	1.788711	1.799192	1.767930	1.745537	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on logical+100x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	0.023101	0.0236420	0.0240043	0.0238795	0.024174	0.027141	100
rowCounts(X,cols,rows)	0.027389	0.0280630	0.0289578	0.0286290	0.029635	0.035786	100
rowCounts(X[cols,rows])	0.039200	0.0398555	0.0407704	0.0400935	0.040480	0.076089	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.185620	1.186998	1.206358	1.198894	1.225904	1.318522	100
rowCounts(X[cols,rows])	1.696896	1.685792	1.698461	1.678992	1.674526	2.803471	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on logical+100x100 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+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	neval
colCounts_X_S	18.947	19.930	20.8104	20.5495	21.263	29.576	100
rowCounts_X_S	23.101	23.642	24.0043	23.8795	24.174	27.141	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
rowCounts_X_S	1.219243	1.186252	1.153476	1.162048	1.136904	0.9176697	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+100x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x10 matrix

> X <- data[["1000x10"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

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

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	0.018640	0.019780	0.0206665	0.0203625	0.021280	0.027603	100
colCounts(X,rows,cols)	0.030097	0.031035	0.0320233	0.0317420	0.032234	0.051039	100
colCounts(X[rows,cols])	0.038484	0.039508	0.0407278	0.0400310	0.040821	0.078229	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.614646	1.569009	1.549524	1.558846	1.514756	1.849038	100
colCounts(X[rows,cols])	2.064592	1.997371	1.970712	1.965918	1.918280	2.834076	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on logical+1000x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	0.027136	0.0280595	0.0291683	0.0284025	0.029076	0.059775	100
rowCounts(X,cols,rows)	0.033165	0.0347540	0.0360522	0.0355060	0.036862	0.054925	100
rowCounts(X[cols,rows])	0.044851	0.0455980	0.0463997	0.0458900	0.047052	0.053661	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
rowCounts(X,cols,rows)	1.222177	1.238582	1.236003	1.250101	1.267781	0.9188624	100
rowCounts(X[cols,rows])	1.652823	1.625047	1.590756	1.615703	1.618242	0.8977164	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on logical+1000x10 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+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	neval
colCounts_X_S	18.640	19.7800	20.66652	20.3625	21.280	27.603	100
rowCounts_X_S	27.136	28.0595	29.16834	28.4025	29.076	59.775	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.455794	1.418579	1.411381	1.394843	1.366353	2.165526	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+1000x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

10x1000 matrix

> X <- data[["10x1000"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

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

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	0.020857	0.022367	0.0231282	0.0228645	0.0234875	0.030590	100
colCounts(X,rows,cols)	0.033442	0.034594	0.0353813	0.0351415	0.0358185	0.042246	100
colCounts(X[rows,cols])	0.037980	0.039530	0.0412870	0.0404555	0.0411230	0.082091	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.603395	1.546654	1.529791	1.536946	1.525003	1.381040	100
colCounts(X[rows,cols])	1.820971	1.767336	1.785137	1.769359	1.750846	2.683589	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on logical+10x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	0.023112	0.0235175	0.0238726	0.023717	0.0240180	0.029028	100
rowCounts(X,cols,rows)	0.027530	0.0286325	0.0294635	0.029170	0.0300180	0.035179	100
rowCounts(X[cols,rows])	0.042246	0.0427420	0.0435779	0.042984	0.0433395	0.077803	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.191156	1.217498	1.234196	1.229920	1.249813	1.211899	100
rowCounts(X[cols,rows])	1.827882	1.817455	1.825437	1.812371	1.804459	2.680274	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on logical+10x1000 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+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	neval
colCounts_X_S	20.857	22.3670	23.12817	22.8645	23.4875	30.590	100
rowCounts_X_S	23.112	23.5175	23.87260	23.7170	24.0180	29.028	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
rowCounts_X_S	1.108117	1.051437	1.032187	1.037285	1.022587	0.9489376	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+10x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x1000 matrix

> X <- data[["100x1000"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

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

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	0.210573	0.2135550	0.2153174	0.2152890	0.2168370	0.223632	100
colCounts(X,rows,cols)	0.267100	0.2696795	0.2726299	0.2705985	0.2715000	0.364316	100
colCounts(X[rows,cols])	0.386361	0.3890185	0.4089667	0.3903310	0.3918915	1.222766	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.268444	1.262811	1.266177	1.256908	1.252093	1.629087	100
colCounts(X[rows,cols])	1.834808	1.821631	1.899366	1.813056	1.807309	5.467759	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on logical+100x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	0.212653	0.2143025	0.2183749	0.2154425	0.2170470	0.275730	100
rowCounts(X,cols,rows)	0.268597	0.2704125	0.2784853	0.2712515	0.2858325	0.327082	100
rowCounts(X[cols,rows])	0.373669	0.3753110	0.3959550	0.3760725	0.3771535	1.189767	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.263076	1.261826	1.275262	1.259044	1.316915	1.186240	100
rowCounts(X[cols,rows])	1.757177	1.751314	1.813189	1.745582	1.737658	4.314971	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on logical+100x1000 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+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	neval
colCounts_X_S	210.573	213.5550	215.3174	215.2890	216.837	223.632	100
rowCounts_X_S	212.653	214.3025	218.3749	215.4425	217.047	275.730	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.0000	1.0000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.009878	1.0035	1.0142	1.000713	1.000968	1.232963	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+100x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x100 matrix

> X <- data[["1000x100"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

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

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	0.206296	0.2088270	0.2105008	0.2095975	0.211003	0.239486	100
colCounts(X,rows,cols)	0.257324	0.2595640	0.2612518	0.2607240	0.261561	0.295740	100
colCounts(X[rows,cols])	0.368593	0.3708345	0.3903450	0.3718035	0.373097	1.183091	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.247353	1.242962	1.241096	1.243927	1.239608	1.234895	100
colCounts(X[rows,cols])	1.786719	1.775798	1.854363	1.773893	1.768207	4.940126	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on logical+1000x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	0.217393	0.218990	0.2234541	0.2204600	0.2218095	0.261642	100
rowCounts(X,cols,rows)	0.277694	0.279795	0.2933039	0.2808940	0.3022375	0.377908	100
rowCounts(X[cols,rows])	0.393659	0.395616	0.4231071	0.3972265	0.4071675	1.205718	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.277382	1.277661	1.312591	1.274127	1.362599	1.444371	100
rowCounts(X[cols,rows])	1.810817	1.806548	1.893485	1.801808	1.835663	4.608274	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on logical+1000x100 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+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	neval
colCounts_X_S	206.296	208.827	210.5009	209.5975	211.0030	239.486	100
rowCounts_X_S	217.393	218.990	223.4541	220.4600	221.8095	261.642	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.053792	1.048667	1.061535	1.051826	1.051215	1.092515	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on logical+1000x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

Data type "integer"

Data

> rmatrix <- function(nrow, ncol, mode = c("logical", "double", "integer", "index"), range = c(-100, 
+     +100), naProb = 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 (naProb > 0) 
+         X[sample(n, size = naProb * 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 matrix

> X <- data[["10x10"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.005416	0.0057005	0.0061513	0.0059020	0.0060645	0.031272	100
colCounts(X,rows,cols)	0.005895	0.0061230	0.0062979	0.0062915	0.0063865	0.007867	100
colCounts(X[rows,cols])	0.007269	0.0075050	0.0077130	0.0076530	0.0077785	0.011217	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
colCounts(X,rows,cols)	1.088442	1.074116	1.023841	1.065995	1.053096	0.2515669	100
colCounts(X[rows,cols])	1.342134	1.316551	1.253885	1.296679	1.282628	0.3586915	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.005269	0.0055080	0.0059156	0.0056255	0.0057990	0.031555	100
rowCounts(X,cols,rows)	0.005696	0.0058985	0.0060822	0.0060715	0.0062050	0.007589	100
rowCounts(X[cols,rows])	0.006934	0.0072965	0.0075060	0.0074625	0.0075995	0.012194	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
rowCounts(X,cols,rows)	1.081040	1.070897	1.028170	1.079282	1.070012	0.2405007	100
rowCounts(X[cols,rows])	1.315999	1.324709	1.268864	1.326549	1.310485	0.3864364	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on integer+10x10 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
2	rowCounts_X_S	5.269	5.5080	5.91555	5.6255	5.7990	31.555	100
1	colCounts_X_S	5.416	5.7005	6.15128	5.9020	6.0645	31.272	100

	expr	min	lq	mean	median	uq	max	neval
2	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
1	colCounts_X_S	1.027899	1.034949	1.039849	1.049151	1.045784	0.9910315	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on integer+10x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x100 matrix

> X <- data[["100x100"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.011864	0.0122845	0.0124659	0.012432	0.0126225	0.014931	100
colCounts(X,rows,cols)	0.013964	0.0144515	0.0149858	0.014630	0.0147830	0.045248	100
colCounts(X[rows,cols])	0.028060	0.0284150	0.0288360	0.028585	0.0287925	0.037297	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.177006	1.176401	1.202144	1.176802	1.171163	3.030473	100
colCounts(X[rows,cols])	2.365138	2.313077	2.313191	2.299308	2.281046	2.497957	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.014489	0.0147995	0.0149601	0.0149240	0.0150840	0.016553	100
rowCounts(X,cols,rows)	0.015342	0.0157365	0.0161278	0.0158935	0.0160595	0.023975	100
rowCounts(X[cols,rows])	0.030468	0.0309350	0.0314921	0.0311505	0.0313750	0.065294	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.058872	1.063313	1.078056	1.064963	1.064671	1.448378	100
rowCounts(X[cols,rows])	2.102837	2.090273	2.105067	2.087276	2.080019	3.944542	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on integer+100x100 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	11.864	12.2845	12.46590	12.432	12.6225	14.931	100
rowCounts_X_S	14.489	14.7995	14.96013	14.924	15.0840	16.553	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.221258	1.20473	1.200084	1.200451	1.195009	1.108633	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on integer+100x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x10 matrix

> X <- data[["1000x10"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.012617	0.0129420	0.0131594	0.013114	0.0133165	0.014948	100
colCounts(X,rows,cols)	0.015865	0.0162025	0.0164124	0.016323	0.0164750	0.021599	100
colCounts(X[rows,cols])	0.034351	0.0348265	0.0354155	0.035082	0.0352675	0.066581	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.257430	1.251932	1.247203	1.244700	1.237187	1.444943	100
colCounts(X[rows,cols])	2.722596	2.690967	2.691273	2.675156	2.648406	4.454174	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.015106	0.015394	0.0158660	0.0155255	0.015647	0.041759	100
rowCounts(X,cols,rows)	0.016699	0.016973	0.0174394	0.0170915	0.017284	0.032095	100
rowCounts(X[cols,rows])	0.032107	0.032626	0.0331494	0.0328140	0.033003	0.042608	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
rowCounts(X,cols,rows)	1.105455	1.102572	1.099165	1.100866	1.104621	0.7685768	100
rowCounts(X[cols,rows])	2.125447	2.119397	2.089327	2.113555	2.109222	1.0203309	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on integer+1000x10 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	12.617	12.942	13.15938	13.1140	13.3165	14.948	100
rowCounts_X_S	15.106	15.394	15.86605	15.5255	15.6470	41.759	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.197274	1.189461	1.205684	1.183887	1.175008	2.793618	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on integer+1000x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

10x1000 matrix

> X <- data[["10x1000"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.012544	0.0128755	0.0130978	0.013028	0.0132255	0.016785	100
colCounts(X,rows,cols)	0.016023	0.0163535	0.0165430	0.016547	0.0167005	0.018115	100
colCounts(X[rows,cols])	0.029667	0.0301705	0.0309225	0.030377	0.0306820	0.065074	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.277344	1.270125	1.263028	1.27011	1.262750	1.079237	100
colCounts(X[rows,cols])	2.365035	2.343249	2.360886	2.33167	2.319912	3.876914	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.014244	0.0146340	0.0149823	0.0147930	0.0149585	0.022779	100
rowCounts(X,cols,rows)	0.015847	0.0161915	0.0163528	0.0163275	0.0164940	0.018114	100
rowCounts(X[cols,rows])	0.033581	0.0339280	0.0346973	0.0340915	0.0343495	0.067027	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
rowCounts(X,cols,rows)	1.112539	1.106430	1.091470	1.103732	1.102651	0.7952061	100
rowCounts(X[cols,rows])	2.357554	2.318436	2.315878	2.304570	2.296320	2.9424909	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on integer+10x1000 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	12.544	12.8755	13.09785	13.028	13.2255	16.785	100
rowCounts_X_S	14.244	14.6340	14.98234	14.793	14.9585	22.779	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.135523	1.136577	1.143878	1.135477	1.131035	1.357105	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on integer+10x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x1000 matrix

> X <- data[["100x1000"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.068132	0.0685380	0.0691408	0.0687405	0.0689565	0.086163	100
colCounts(X,rows,cols)	0.088553	0.0890335	0.0906975	0.0892770	0.0895445	0.181805	100
colCounts(X[rows,cols])	0.243007	0.2442635	0.2623006	0.2447515	0.2453595	1.067442	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.299727	1.299038	1.311779	1.298754	1.298565	2.110012	100
colCounts(X[rows,cols])	3.566709	3.563913	3.793715	3.560514	3.558178	12.388635	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.091334	0.0918785	0.0932175	0.0921075	0.0924055	0.117729	100
rowCounts(X,cols,rows)	0.094401	0.0948775	0.0968161	0.0950850	0.0955015	0.127119	100
rowCounts(X[cols,rows])	0.251592	0.2528315	0.2755978	0.2533825	0.2556230	1.137349	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.033580	1.032641	1.038604	1.032326	1.033505	1.079759	100
rowCounts(X[cols,rows])	2.754637	2.751803	2.956502	2.750943	2.766318	9.660738	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on integer+100x1000 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	68.132	68.5380	69.14082	68.7405	68.9565	86.163	100
rowCounts_X_S	91.334	91.8785	93.21751	92.1075	92.4055	117.729	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.340545	1.340548	1.348227	1.339931	1.340055	1.366352	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on integer+100x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x100 matrix

> X <- data[["1000x100"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.063982	0.0644120	0.0652741	0.0646515	0.0648415	0.091861	100
colCounts(X,rows,cols)	0.081098	0.0814875	0.0825432	0.0816145	0.0818045	0.116981	100
colCounts(X[rows,cols])	0.225304	0.2266645	0.2437216	0.2272195	0.2277855	1.050614	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.267513	1.265098	1.264563	1.262376	1.261607	1.273457	100
colCounts(X[rows,cols])	3.521365	3.518979	3.733819	3.514528	3.512958	11.436997	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.095760	0.0962295	0.0969354	0.0964120	0.0966525	0.126040	100
rowCounts(X,cols,rows)	0.101907	0.1023455	0.1038735	0.1025330	0.1029950	0.209546	100
rowCounts(X[cols,rows])	0.272514	0.2738145	0.2913992	0.2744165	0.2751320	1.088787	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.064192	1.063556	1.071574	1.063488	1.065622	1.662536	100
rowCounts(X[cols,rows])	2.845802	2.845432	3.006118	2.846290	2.846610	8.638424	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on integer+1000x100 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	63.982	64.4120	65.27407	64.6515	64.8415	91.861	100
rowCounts_X_S	95.760	96.2295	96.93537	96.4120	96.6525	126.040	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.496671	1.493969	1.485052	1.491257	1.490596	1.372073	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_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), naProb = 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 (naProb > 0) 
+         X[sample(n, size = naProb * 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 matrix

> X <- data[["10x10"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.005776	0.0060265	0.0065510	0.0062320	0.0063900	0.034787	100
colCounts(X,rows,cols)	0.006246	0.0064510	0.0067978	0.0066205	0.0067825	0.015304	100
colCounts(X[rows,cols])	0.007646	0.0079705	0.0082681	0.0081205	0.0083660	0.012421	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
colCounts(X,rows,cols)	1.081371	1.070439	1.037670	1.062339	1.061424	0.4399345	100
colCounts(X[rows,cols])	1.323754	1.322575	1.262102	1.303033	1.309233	0.3570587	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.005661	0.0059185	0.0063226	0.0060510	0.0062015	0.030947	100
rowCounts(X,cols,rows)	0.006097	0.0063860	0.0066006	0.0065365	0.0066595	0.009885	100
rowCounts(X[cols,rows])	0.007422	0.0077140	0.0079681	0.0079370	0.0080815	0.012554	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
rowCounts(X,cols,rows)	1.077018	1.078990	1.043979	1.080235	1.073853	0.3194171	100
rowCounts(X[cols,rows])	1.311076	1.303371	1.260251	1.311684	1.303152	0.4056613	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on double+10x10 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
2	rowCounts_X_S	5.661	5.9185	6.32259	6.051	6.2015	30.947	100
1	colCounts_X_S	5.776	6.0265	6.55103	6.232	6.3900	34.787	100

	expr	min	lq	mean	median	uq	max	neval
2	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
1	colCounts_X_S	1.020314	1.018248	1.036131	1.029912	1.030396	1.124083	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on double+10x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x100 matrix

> X <- data[["100x100"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.022840	0.023279	0.0234452	0.0233915	0.023573	0.025221	100
colCounts(X,rows,cols)	0.024030	0.024287	0.0248283	0.0244435	0.024567	0.059807	100
colCounts(X[rows,cols])	0.040091	0.040496	0.0407492	0.0406490	0.040848	0.044909	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.052102	1.043301	1.058993	1.044974	1.042167	2.371318	100
colCounts(X[rows,cols])	1.755298	1.739594	1.738062	1.737768	1.732830	1.780619	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.028664	0.0289325	0.0290869	0.0290480	0.029165	0.031082	100
rowCounts(X,cols,rows)	0.029365	0.0297075	0.0300238	0.0298345	0.030020	0.035005	100
rowCounts(X[cols,rows])	0.046469	0.0468455	0.0475393	0.0471835	0.047487	0.083667	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.024456	1.026787	1.032210	1.027076	1.029316	1.126214	100
rowCounts(X[cols,rows])	1.621162	1.619131	1.634387	1.624329	1.628219	2.691815	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on double+100x100 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	22.840	23.2790	23.44519	23.3915	23.573	25.221	100
rowCounts_X_S	28.664	28.9325	29.08693	29.0480	29.165	31.082	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.254991	1.242858	1.240635	1.241819	1.237221	1.232386	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on double+100x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x10 matrix

> X <- data[["1000x10"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.022194	0.0225290	0.0233550	0.0226715	0.0229025	0.054181	100
colCounts(X,rows,cols)	0.023808	0.0241005	0.0250075	0.0242805	0.0245185	0.054084	100
colCounts(X[rows,cols])	0.042277	0.0429260	0.0445920	0.0431880	0.0435835	0.080696	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
colCounts(X,rows,cols)	1.072722	1.069754	1.070753	1.070970	1.070560	0.9982097	100
colCounts(X[rows,cols])	1.904884	1.905366	1.909308	1.904947	1.903002	1.4893782	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.029197	0.0297760	0.0318996	0.0303520	0.0314330	0.062679	100
rowCounts(X,cols,rows)	0.031702	0.0322705	0.0341869	0.0331555	0.0340710	0.054385	100
rowCounts(X[cols,rows])	0.050092	0.0514180	0.0548176	0.0529750	0.0543675	0.082837	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.085797	1.083776	1.071700	1.092366	1.083924	0.867675	100
rowCounts(X[cols,rows])	1.715656	1.726827	1.718441	1.745354	1.729631	1.321607	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on double+1000x10 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	22.194	22.529	23.35505	22.6715	22.9025	54.181	100
rowCounts_X_S	29.197	29.776	31.89964	30.3520	31.4330	62.679	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000	100
rowCounts_X_S	1.315536	1.321674	1.365856	1.338773	1.37247	1.156845	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on double+1000x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

10x1000 matrix

> X <- data[["10x1000"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.023795	0.0241870	0.0244465	0.0244075	0.024690	0.027255	100
colCounts(X,rows,cols)	0.027009	0.0273575	0.0278870	0.0275005	0.027737	0.047748	100
colCounts(X[rows,cols])	0.043498	0.0444555	0.0455218	0.0447985	0.045130	0.082122	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.135070	1.131083	1.140735	1.126723	1.123410	1.751899	100
colCounts(X[rows,cols])	1.828031	1.837992	1.862096	1.835440	1.827865	3.013098	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.026702	0.027090	0.0274589	0.0272360	0.0274380	0.037999	100
rowCounts(X,cols,rows)	0.028191	0.028617	0.0289895	0.0287405	0.0289060	0.042267	100
rowCounts(X[cols,rows])	0.046817	0.047344	0.0482336	0.0475515	0.0478495	0.085224	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.055764	1.056368	1.055740	1.055239	1.053502	1.112319	100
rowCounts(X[cols,rows])	1.753314	1.747656	1.756576	1.745906	1.743914	2.242796	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on double+10x1000 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	23.795	24.187	24.44652	24.4075	24.690	27.255	100
rowCounts_X_S	26.702	27.090	27.45889	27.2360	27.438	37.999	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.0000	1.000000	100
rowCounts_X_S	1.122168	1.120023	1.123223	1.115887	1.1113	1.394203	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on double+10x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x1000 matrix

> X <- data[["100x1000"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.176507	0.1770745	0.1824289	0.1774740	0.1821225	0.229395	100
colCounts(X,rows,cols)	0.183405	0.1839240	0.1926547	0.1843275	0.1914855	0.344075	100
colCounts(X[rows,cols])	0.362784	0.3646015	0.4347960	0.3666355	0.3951210	1.290293	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.039081	1.038682	1.056054	1.038617	1.051411	1.499924	100
colCounts(X[rows,cols])	2.055352	2.059029	2.383373	2.065855	2.169534	5.624765	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.213232	0.2137795	0.2172787	0.214029	0.2143360	0.296912	100
rowCounts(X,cols,rows)	0.215175	0.2158875	0.2187254	0.216126	0.2163690	0.283987	100
rowCounts(X[cols,rows])	0.391095	0.3973470	0.4712330	0.399968	0.4033565	1.213951	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000	100
rowCounts(X,cols,rows)	1.009112	1.009861	1.006658	1.009798	1.009485	0.9564686	100
rowCounts(X[cols,rows])	1.834129	1.858677	2.168795	1.868756	1.881889	4.0885885	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on double+100x1000 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	176.507	177.0745	182.4289	177.474	182.1225	229.395	100
rowCounts_X_S	213.232	213.7795	217.2787	214.029	214.3360	296.912	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.208065	1.207286	1.191032	1.205974	1.176878	1.294326	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on double+100x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x100 matrix

> X <- data[["1000x100"]]
> rows <- sample.int(dim(X)[1], size = dim(X)[1] * 0.7)
> cols <- sample.int(dim(X)[2], size = dim(X)[2] * 0.7)
> X_S <- X[rows, cols]
> value <- 42

> colStats <- microbenchmark(colCounts_X_S = colCounts(X_S, value = value, na.rm = FALSE), `colCounts(X,rows,cols)` = colCounts(X, 
+     value = value, na.rm = FALSE, rows = rows, cols = cols), `colCounts(X[rows,cols])` = colCounts(X[rows, 
+     cols], value = value, na.rm = FALSE), unit = "ms")

> X <- t(X)
> X_S <- t(X_S)

> rowStats <- microbenchmark(rowCounts_X_S = rowCounts(X_S, value = value, na.rm = FALSE), `rowCounts(X,cols,rows)` = rowCounts(X, 
+     value = value, na.rm = FALSE, rows = cols, cols = rows), `rowCounts(X[cols,rows])` = rowCounts(X[cols, 
+     rows], value = value, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(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	neval
colCounts_X_S	0.172232	0.1727555	0.1761024	0.1730220	0.1775325	0.204671	100
colCounts(X,rows,cols)	0.176541	0.1769315	0.1809258	0.1772205	0.1819450	0.255468	100
colCounts(X[rows,cols])	0.356451	0.3595605	0.4228386	0.3610030	0.3692855	1.199084	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
colCounts(X,rows,cols)	1.025019	1.024173	1.027390	1.024266	1.024855	1.248189	100
colCounts(X[rows,cols])	2.069598	2.081326	2.401095	2.086457	2.080101	5.858593	100
Table: Benchmarking of rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(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	neval
rowCounts_X_S	0.217591	0.2181705	0.2216322	0.2183865	0.2192595	0.269604	100
rowCounts(X,cols,rows)	0.219859	0.2203595	0.2254859	0.2207110	0.2236775	0.410901	100
rowCounts(X[cols,rows])	0.402089	0.4058085	0.8574803	0.4078440	0.4189660	39.273514	100

expr	min	lq	mean	median	uq	max	neval
rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts(X,cols,rows)	1.010423	1.010033	1.017388	1.010644	1.020150	1.524091	100
rowCounts(X[cols,rows])	1.847912	1.860052	3.868934	1.867533	1.910823	145.671110	100
Figure: Benchmarking of colCounts_X_S(), colCounts(X,rows,cols)() and colCounts(X[rows,cols])() on double+1000x100 data as well as rowCounts_X_S(), rowCounts(X,cols,rows)() and rowCounts(X[cols,rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colCounts_X_S() and rowCounts_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	neval
colCounts_X_S	172.232	172.7555	176.1024	173.0220	177.5325	204.671	100
rowCounts_X_S	217.591	218.1705	221.6322	218.3865	219.2595	269.604	100

expr	min	lq	mean	median	uq	max	neval
colCounts_X_S	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000	100
rowCounts_X_S	1.26336	1.262886	1.258542	1.262189	1.235039	1.317255	100
Figure: Benchmarking of colCounts_X_S() and rowCounts_X_S() on double+1000x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

Appendix

Session information

R version 3.0.2 (2013-09-25)
Platform: x86_64-pc-linux-gnu (64-bit)

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_1.4-2 matrixStats_0.14.0   ggplot2_1.0.1       
[4] knitr_1.10.5         R.devices_2.13.0     R.utils_2.0.2       
[7] R.oo_1.19.0          R.methodsS3_1.7.0   

loaded via a namespace (and not attached):
 [1] base64enc_0.1-2    colorspace_1.2-6   dichromat_2.0-0    digest_0.6.8      
 [5] grid_3.0.2         gtable_0.1.2       highr_0.5          labeling_0.3      
 [9] magrittr_1.5       MASS_7.3-29        munsell_0.4.2      plyr_1.8          
[13] proto_0.3-10       R.cache_0.10.0     RColorBrewer_1.1-2 reshape2_1.2.2    
[17] R.rsp_0.20.0       scales_0.2.3       stringi_0.4-1      stringr_1.0.0     
[21] tools_3.0.2       

Total processing time was 14.81 secs.

Reproducibility

To reproduce this report, do:

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

---

Copyright Dongcan Jiang. Last updated on 2015-05-09 07:47:48 (+0000 UTC). Powered by RSP.

<script> var link = document.createElement('link'); link.rel = 'icon'; link.href = "data:image/png;base64,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" document.getElementsByTagName('head')[0].appendChild(link); </script>