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), 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 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]
> 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
1	colCounts_X_S	0.002247	0.002308	0.0043100	0.0023660	0.002528	0.190819
2	colCounts(X, rows, cols)	0.002479	0.002549	0.0026707	0.0026115	0.002716	0.004846
3	colCounts(X[rows, cols])	0.003019	0.003260	0.0034636	0.0033335	0.003463	0.010424

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	colCounts(X, rows, cols)	1.103249	1.104419	0.6196561	1.103762	1.074367	0.0253958
3	colCounts(X[rows, cols])	1.343569	1.412478	0.8036158	1.408918	1.369858	0.0546277

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
1	rowCounts_X_S	0.002144	0.002213	0.0023419	0.0022610	0.0024285	0.004592
2	rowCounts(X, cols, rows)	0.002339	0.002460	0.0045045	0.0025515	0.0026755	0.194337
3	rowCounts(X[cols, rows])	0.002803	0.002999	0.0032171	0.0030970	0.0032425	0.007988

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.090952	1.111613	1.923423	1.128483	1.101709	42.320775
3	rowCounts(X[cols, rows])	1.307369	1.355174	1.373686	1.369748	1.335186	1.739547

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
2	rowCounts_X_S	2.144	2.213	2.34194	2.261	2.4285	4.592
1	colCounts_X_S	2.247	2.308	4.31002	2.366	2.5280	190.819

	expr	min	lq	mean	median	uq	max
2	rowCounts_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.00000
1	colCounts_X_S	1.048041	1.042928	1.840363	1.04644	1.040972	41.55466

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.014956	0.0159250	0.0187822	0.0163800	0.0170005	0.065345
2	colCounts(X, rows, cols)	0.026573	0.0277135	0.0301256	0.0280595	0.0286710	0.064991
3	colCounts(X[rows, cols])	0.030900	0.0321415	0.0361838	0.0325290	0.0334035	0.094142

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colCounts(X, rows, cols)	1.776745	1.740251	1.603940	1.713034	1.686480	0.9945826
3	colCounts(X[rows, cols])	2.066060	2.018305	1.926489	1.985897	1.964854	1.4406917

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
1	rowCounts_X_S	0.018748	0.0199265	0.0204167	0.0202775	0.0208110	0.024813
2	rowCounts(X, cols, rows)	0.021332	0.0246285	0.0255354	0.0257890	0.0265035	0.029800
3	rowCounts(X[cols, rows])	0.026049	0.0273755	0.0285468	0.0279035	0.0284620	0.085041

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.137828	1.235967	1.250716	1.271804	1.273533	1.200983
3	rowCounts(X[cols, rows])	1.389428	1.373824	1.398210	1.376082	1.367642	3.427276

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
1	colCounts_X_S	14.956	15.9250	18.78225	16.3800	17.0005	65.345
2	rowCounts_X_S	18.748	19.9265	20.41665	20.2775	20.8110	24.813

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts_X_S	1.253544	1.251272	1.087018	1.237943	1.224141	0.379723

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.014952	0.0155295	0.0159807	0.0159125	0.0162535	0.023835
2	colCounts(X, rows, cols)	0.026870	0.0275695	0.0282494	0.0278060	0.0281060	0.066585
3	colCounts(X[rows, cols])	0.030305	0.0312165	0.0318215	0.0315610	0.0319885	0.045172

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.797084	1.775299	1.767713	1.747431	1.729228	2.793581
3	colCounts(X[rows, cols])	2.026819	2.010142	1.991242	1.983409	1.968099	1.895196

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
1	rowCounts_X_S	0.018717	0.0200250	0.0210584	0.0206775	0.0211855	0.063188
2	rowCounts(X, cols, rows)	0.023208	0.0262735	0.0274096	0.0273810	0.0283210	0.043252
3	rowCounts(X[cols, rows])	0.028331	0.0292615	0.0301990	0.0298515	0.0304655	0.043113

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCounts(X, cols, rows)	1.239942	1.312035	1.301603	1.324193	1.336811	0.6844971
3	rowCounts(X[cols, rows])	1.513651	1.461248	1.434062	1.443671	1.438035	0.6822973

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
1	colCounts_X_S	14.952	15.5295	15.98073	15.9125	16.2535	23.835
2	rowCounts_X_S	18.717	20.0250	21.05837	20.6775	21.1855	63.188

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000
2	rowCounts_X_S	1.251806	1.289481	1.317735	1.29945	1.303442	2.651059

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.016053	0.0173490	0.0181913	0.0180045	0.0187555	0.026057
3	colCounts(X[rows, cols])	0.025500	0.0266630	0.0275438	0.0272840	0.0281125	0.037193
2	colCounts(X, rows, cols)	0.031713	0.0331705	0.0340044	0.0336375	0.0341080	0.071593

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colCounts(X[rows, cols])	1.588488	1.536861	1.514123	1.515399	1.498894	1.427371
2	colCounts(X, rows, cols)	1.975519	1.911955	1.869269	1.868283	1.818560	2.747553

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
1	rowCounts_X_S	0.018273	0.0194185	0.0206247	0.0202325	0.0207075	0.062510
2	rowCounts(X, cols, rows)	0.022733	0.0253140	0.0264121	0.0265370	0.0277050	0.030938
3	rowCounts(X[cols, rows])	0.026197	0.0276300	0.0285089	0.0282855	0.0287930	0.045801

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCounts(X, cols, rows)	1.244076	1.303602	1.280602	1.311603	1.337921	0.4949288
3	rowCounts(X[cols, rows])	1.433645	1.422870	1.382266	1.398023	1.390462	0.7326988

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
1	colCounts_X_S	16.053	17.3490	18.19127	18.0045	18.7555	26.057
2	rowCounts_X_S	18.273	19.4185	20.62474	20.2325	20.7075	62.510

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts_X_S	1.138292	1.119286	1.133771	1.123747	1.104076	2.398972

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.177938	0.182584	0.1854838	0.1868145	0.1888510	0.200763
2	colCounts(X, rows, cols)	0.253978	0.257596	0.2662256	0.2648400	0.2708645	0.356632
3	colCounts(X[rows, cols])	0.324466	0.328664	0.3373513	0.3360885	0.3437515	0.365653

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.427340	1.410836	1.435304	1.417663	1.434276	1.776383
3	colCounts(X[rows, cols])	1.823478	1.800070	1.818764	1.799049	1.820226	1.821317

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
1	rowCounts_X_S	0.181591	0.1961155	0.2004233	0.2013035	0.2037485	0.250633
2	rowCounts(X, cols, rows)	0.239456	0.2551820	0.2620845	0.2624635	0.2682000	0.301425
3	rowCounts(X[cols, rows])	0.242631	0.2628810	0.2689877	0.2650875	0.2720390	0.329904

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.318656	1.301182	1.307655	1.303820	1.316329	1.202655
3	rowCounts(X[cols, rows])	1.336140	1.340440	1.342098	1.316855	1.335171	1.316283

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
1	colCounts_X_S	177.938	182.5840	185.4838	186.8145	188.8510	200.763
2	rowCounts_X_S	181.591	196.1155	200.4233	201.3035	203.7485	250.633

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts_X_S	1.02053	1.074111	1.080543	1.077558	1.078885	1.248402

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.172271	0.1736790	0.1772729	0.1766780	0.1788080	0.195033
3	colCounts(X[rows, cols])	0.235371	0.2379725	0.2439298	0.2429765	0.2451605	0.321591
2	colCounts(X, rows, cols)	0.243631	0.2462140	0.2507330	0.2492495	0.2536165	0.269588

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colCounts(X[rows, cols])	1.366283	1.370186	1.376013	1.375250	1.371082	1.648906
2	colCounts(X, rows, cols)	1.414231	1.417638	1.414390	1.410756	1.418373	1.382269

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
1	rowCounts_X_S	0.189107	0.1913065	0.2006118	0.200190	0.2016930	0.356354
2	rowCounts(X, cols, rows)	0.252300	0.2594665	0.2675861	0.266562	0.2698795	0.478265
3	rowCounts(X[cols, rows])	0.254775	0.2587900	0.2753844	0.269315	0.2737910	0.705173

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.334165	1.356287	1.333851	1.331545	1.338071	1.342106
3	rowCounts(X[cols, rows])	1.347253	1.352751	1.372723	1.345297	1.357464	1.978855

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
1	colCounts_X_S	172.271	173.6790	177.2729	176.678	178.808	195.033
2	rowCounts_X_S	189.107	191.3065	200.6118	200.190	201.693	356.354

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts_X_S	1.09773	1.101495	1.131655	1.133078	1.127986	1.827147

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), 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 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]
> 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
1	colCounts_X_S	0.002144	0.0022385	0.0028208	0.0023435	0.0025190	0.041531
2	colCounts(X, rows, cols)	0.002373	0.0024670	0.0026433	0.0025505	0.0026750	0.004961
3	colCounts(X[rows, cols])	0.002873	0.0031260	0.0034178	0.0032445	0.0034305	0.010647

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	colCounts(X, rows, cols)	1.106810	1.102077	0.9370916	1.088329	1.061929	0.1194529
3	colCounts(X[rows, cols])	1.340019	1.396471	1.2116443	1.384468	1.361850	0.2563627

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
1	rowCounts_X_S	0.002096	0.0021575	0.0022643	0.002203	0.0023065	0.004211
2	rowCounts(X, cols, rows)	0.002327	0.0023905	0.0028323	0.002459	0.0026100	0.034085
3	rowCounts(X[cols, rows])	0.002748	0.0029110	0.0031089	0.003015	0.0031515	0.007511

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.110210	1.107995	1.250843	1.116205	1.131585	8.094277
3	rowCounts(X[cols, rows])	1.311069	1.349247	1.372977	1.368588	1.366356	1.783662

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
2	rowCounts_X_S	2.096	2.1575	2.26432	2.2030	2.3065	4.211
1	colCounts_X_S	2.144	2.2385	2.82077	2.3435	2.5190	41.531

	expr	min	lq	mean	median	uq	max
2	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCounts_X_S	1.022901	1.037543	1.245747	1.063777	1.092131	9.862503

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.005496	0.005886	0.0060842	0.0059955	0.0061705	0.010629
2	colCounts(X, rows, cols)	0.009536	0.009996	0.0102940	0.0101595	0.0103535	0.013073
3	colCounts(X[rows, cols])	0.013371	0.013768	0.0146739	0.0139555	0.0141520	0.073751

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.73508	1.698267	1.691933	1.694521	1.677903	1.229937
3	colCounts(X[rows, cols])	2.43286	2.339110	2.411813	2.327662	2.293493	6.938658

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
1	rowCounts_X_S	0.006290	0.0064845	0.0066874	0.0065965	0.0067285	0.010290
2	rowCounts(X, cols, rows)	0.008769	0.0090920	0.0092881	0.0092510	0.0094130	0.010694
3	rowCounts(X[cols, rows])	0.014183	0.0143920	0.0152446	0.0145455	0.0147010	0.072689

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.394118	1.402113	1.388899	1.402410	1.398975	1.039261
3	rowCounts(X[cols, rows])	2.254849	2.219446	2.279622	2.205033	2.184885	7.064043

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
1	colCounts_X_S	5.496	5.8860	6.08419	5.9955	6.1705	10.629
2	rowCounts_X_S	6.290	6.4845	6.68736	6.5965	6.7285	10.290

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.00000	1.0000000
2	rowCounts_X_S	1.144469	1.101682	1.099137	1.100242	1.09043	0.9681061

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.005173	0.0053310	0.0054957	0.0054410	0.005562	0.009227
2	colCounts(X, rows, cols)	0.009233	0.0094115	0.0099716	0.0094985	0.009600	0.054607
3	colCounts(X[rows, cols])	0.012814	0.0131405	0.0136673	0.0132700	0.013420	0.026789

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.784844	1.765429	1.814454	1.745727	1.725998	5.918175
3	colCounts(X[rows, cols])	2.477093	2.464922	2.486926	2.438890	2.412801	2.903327

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
1	rowCounts_X_S	0.006286	0.0064400	0.0069763	0.006545	0.006684	0.046441
2	rowCounts(X, cols, rows)	0.009576	0.0098110	0.0101216	0.009976	0.010136	0.023467
3	rowCounts(X[cols, rows])	0.014943	0.0152445	0.0157533	0.015379	0.015555	0.026122

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCounts(X, cols, rows)	1.523385	1.523447	1.450855	1.524217	1.516457	0.5053078
3	rowCounts(X[cols, rows])	2.377187	2.367158	2.258122	2.349733	2.327199	0.5624771

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
1	colCounts_X_S	5.173	5.331	5.49566	5.441	5.562	9.227
2	rowCounts_X_S	6.286	6.440	6.97630	6.545	6.684	46.441

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.00000	1.000000	1.000000	1.000000
2	rowCounts_X_S	1.215156	1.208028	1.26942	1.202904	1.201726	5.033163

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.007496	0.0078670	0.0080659	0.0079735	0.0081920	0.010714
2	colCounts(X, rows, cols)	0.011532	0.0117125	0.0124047	0.0118890	0.0120270	0.062186
3	colCounts(X[rows, cols])	0.016238	0.0167235	0.0171649	0.0168835	0.0171365	0.027109

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.538420	1.488814	1.537918	1.491064	1.468140	5.804181
3	colCounts(X[rows, cols])	2.166222	2.125779	2.128078	2.117452	2.091858	2.530241

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
1	rowCounts_X_S	0.005678	0.0059825	0.0063894	0.0061180	0.006240	0.031608
2	rowCounts(X, cols, rows)	0.009422	0.0098095	0.0100066	0.0099505	0.010155	0.011903
3	rowCounts(X[cols, rows])	0.013411	0.0137055	0.0141973	0.0138955	0.014129	0.027171

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCounts(X, cols, rows)	1.659387	1.639699	1.566125	1.626430	1.627404	0.3765819
3	rowCounts(X[cols, rows])	2.361923	2.290932	2.222014	2.271249	2.264263	0.8596241

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
2	rowCounts_X_S	5.678	5.9825	6.38938	6.1180	6.240	31.608
1	colCounts_X_S	7.496	7.8670	8.06592	7.9735	8.192	10.714

	expr	min	lq	mean	median	uq	max
2	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.00000	1.0000000
1	colCounts_X_S	1.320183	1.315002	1.262395	1.303285	1.31282	0.3389648

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.034451	0.0356035	0.0371499	0.0367555	0.0372390	0.063631
2	colCounts(X, rows, cols)	0.075804	0.0766235	0.0808626	0.0800180	0.0808165	0.190477
3	colCounts(X[rows, cols])	0.106091	0.1082505	0.1128680	0.1122005	0.1150135	0.161123

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	2.200343	2.152134	2.176658	2.177035	2.170211	2.993462
3	colCounts(X[rows, cols])	3.079475	3.040446	3.038177	3.052618	3.088523	2.532146

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
1	rowCounts_X_S	0.033385	0.0344460	0.0365511	0.0357950	0.0375155	0.077913
2	rowCounts(X, cols, rows)	0.066724	0.0673230	0.0706222	0.0693905	0.0733705	0.082511
3	rowCounts(X[cols, rows])	0.096389	0.0993205	0.1031130	0.1018550	0.1053390	0.162200

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.998622	1.954450	1.932152	1.938553	1.955738	1.059014
3	rowCounts(X[cols, rows])	2.887195	2.883368	2.821067	2.845509	2.807879	2.081809

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
2	rowCounts_X_S	33.385	34.4460	36.55107	35.7950	37.5155	77.913
1	colCounts_X_S	34.451	35.6035	37.14990	36.7555	37.2390	63.631

	expr	min	lq	mean	median	uq	max
2	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.0000000	1.000000
1	colCounts_X_S	1.031931	1.033603	1.016383	1.026833	0.9926297	0.816693

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.031942	0.0327945	0.0339031	0.0337160	0.034815	0.040094
2	colCounts(X, rows, cols)	0.067257	0.0678565	0.0705423	0.0696815	0.071716	0.086996
3	colCounts(X[rows, cols])	0.095263	0.0966255	0.1000471	0.0987760	0.100986	0.169547

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	2.105598	2.069143	2.080705	2.066719	2.059917	2.169801
3	colCounts(X[rows, cols])	2.982374	2.946394	2.950976	2.929648	2.900646	4.228738

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
1	rowCounts_X_S	0.042705	0.0433525	0.0460010	0.0458090	0.0467780	0.061322
2	rowCounts(X, cols, rows)	0.069610	0.0715705	0.0745829	0.0740770	0.0761745	0.095900
3	rowCounts(X[cols, rows])	0.113582	0.1163195	0.1221983	0.1200425	0.1258025	0.221970

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.630020	1.650897	1.621333	1.617084	1.628426	1.563876
3	rowCounts(X[cols, rows])	2.659689	2.683109	2.656428	2.620500	2.689352	3.619745

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
1	colCounts_X_S	31.942	32.7945	33.90306	33.716	34.815	40.094
2	rowCounts_X_S	42.705	43.3525	46.00098	45.809	46.778	61.322

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts_X_S	1.336955	1.321944	1.356839	1.358672	1.343616	1.529456

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), 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 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]
> 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
1	colCounts_X_S	0.002166	0.0022495	0.0025563	0.0023085	0.0024025	0.022475
2	colCounts(X, rows, cols)	0.002423	0.0024690	0.0025944	0.0025490	0.0026380	0.004998
3	colCounts(X[rows, cols])	0.003008	0.0032495	0.0034260	0.0033440	0.0034830	0.007832

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.0000000
2	colCounts(X, rows, cols)	1.118652	1.097577	1.014912	1.10418	1.098023	0.2223804
3	colCounts(X[rows, cols])	1.388735	1.444543	1.340199	1.44856	1.449740	0.3484761

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
1	rowCounts_X_S	0.002080	0.002135	0.0024487	0.0021925	0.0023310	0.020384
2	rowCounts(X, cols, rows)	0.002277	0.002364	0.0029680	0.0024320	0.0026310	0.044465
3	rowCounts(X[cols, rows])	0.002776	0.003009	0.0034619	0.0030740	0.0032455	0.028448

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.094712	1.107260	1.212078	1.109236	1.128700	2.181368
3	rowCounts(X[cols, rows])	1.334615	1.409368	1.413800	1.402052	1.392321	1.395604

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
2	rowCounts_X_S	2.080	2.1350	2.44867	2.1925	2.3310	20.384
1	colCounts_X_S	2.166	2.2495	2.55630	2.3085	2.4025	22.475

	expr	min	lq	mean	median	uq	max
2	rowCounts_X_S	1.000000	1.00000	1.000000	1.000000	1.000000	1.00000
1	colCounts_X_S	1.041346	1.05363	1.043954	1.052908	1.030674	1.10258

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.008043	0.008513	0.0102419	0.0086815	0.0109740	0.020431
2	colCounts(X, rows, cols)	0.011426	0.011750	0.0159653	0.0120705	0.0213945	0.035573
3	colCounts(X[rows, cols])	0.026096	0.026396	0.0324221	0.0265765	0.0427830	0.110035

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.420614	1.380242	1.558823	1.39037	1.949563	1.741129
3	colCounts(X[rows, cols])	3.244560	3.100670	3.165623	3.06128	3.898579	5.385688

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
1	rowCounts_X_S	0.008175	0.0083980	0.0088239	0.0085045	0.0086350	0.015719
2	rowCounts(X, cols, rows)	0.011173	0.0115305	0.0119324	0.0116870	0.0118755	0.021638
3	rowCounts(X[cols, rows])	0.016311	0.0166865	0.0180458	0.0168805	0.0170895	0.065964

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.366728	1.373006	1.352283	1.374214	1.375275	1.376551
3	rowCounts(X[cols, rows])	1.995229	1.986961	2.045103	1.984890	1.979097	4.196450

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
2	rowCounts_X_S	8.175	8.398	8.82389	8.5045	8.635	15.719
1	colCounts_X_S	8.043	8.513	10.24192	8.6815	10.974	20.431

	expr	min	lq	mean	median	uq	max
2	rowCounts_X_S	1.0000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCounts_X_S	0.9838532	1.013694	1.160704	1.020812	1.270874	1.299765

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.006852	0.0070605	0.0072335	0.007179	0.0072950	0.011488
2	colCounts(X, rows, cols)	0.010001	0.0102500	0.0109706	0.010395	0.0105430	0.062343
3	colCounts(X[rows, cols])	0.022905	0.0231390	0.0234086	0.023244	0.0234055	0.033792

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.459574	1.451739	1.516646	1.447973	1.445237	5.426793
3	colCounts(X[rows, cols])	3.342820	3.277247	3.236155	3.237777	3.208430	2.941504

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
1	rowCounts_X_S	0.007731	0.0079720	0.0103602	0.0080715	0.0083555	0.047387
2	rowCounts(X, cols, rows)	0.011249	0.0116270	0.0137319	0.0119290	0.0121415	0.039464
3	rowCounts(X[cols, rows])	0.016894	0.0174715	0.0235917	0.0176865	0.0179330	0.109830

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCounts(X, cols, rows)	1.455051	1.458480	1.325450	1.477916	1.453115	0.8328022
3	rowCounts(X[cols, rows])	2.185228	2.191608	2.277161	2.191228	2.146251	2.3177243

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
1	colCounts_X_S	6.852	7.0605	7.23347	7.1790	7.2950	11.488
2	rowCounts_X_S	7.731	7.9720	10.36016	8.0715	8.3555	47.387

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts_X_S	1.128284	1.129099	1.432253	1.124321	1.145374	4.124913

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.008270	0.008571	0.0088871	0.0087415	0.0089455	0.012358
2	colCounts(X, rows, cols)	0.012117	0.012465	0.0133405	0.0126575	0.0128990	0.071094
3	colCounts(X[rows, cols])	0.017551	0.018219	0.0185476	0.0183810	0.0187755	0.024144

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.465175	1.454323	1.501099	1.447978	1.441954	5.752873
3	colCounts(X[rows, cols])	2.122249	2.125656	2.087013	2.102728	2.098876	1.953714

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
1	rowCounts_X_S	0.007726	0.0081005	0.0088008	0.0082590	0.0084790	0.048638
2	rowCounts(X, cols, rows)	0.011144	0.0114290	0.0117761	0.0115640	0.0118250	0.015960
3	rowCounts(X[cols, rows])	0.015795	0.0162430	0.0166161	0.0164035	0.0166565	0.033213

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCounts(X, cols, rows)	1.442402	1.410901	1.338065	1.400170	1.394622	0.3281385
3	rowCounts(X[cols, rows])	2.044395	2.005185	1.888019	1.986136	1.964442	0.6828611

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
2	rowCounts_X_S	7.726	8.1005	8.80081	8.2590	8.4790	48.638
1	colCounts_X_S	8.270	8.5710	8.88714	8.7415	8.9455	12.358

	expr	min	lq	mean	median	uq	max
2	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colCounts_X_S	1.070412	1.058083	1.009809	1.058421	1.055018	0.2540812

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.049440	0.0524380	0.0539035	0.0539400	0.055366	0.067887
2	colCounts(X, rows, cols)	0.078341	0.0811430	0.0855115	0.0851465	0.087653	0.179032
3	colCounts(X[rows, cols])	0.196544	0.2037805	0.2146001	0.2149090	0.222064	0.244480

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.584567	1.547408	1.586381	1.578541	1.583156	2.637206
3	colCounts(X[rows, cols])	3.975404	3.886123	3.981191	3.984223	4.010837	3.601279

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
1	rowCounts_X_S	0.054829	0.0577135	0.0612434	0.061151	0.062375	0.105587
2	rowCounts(X, cols, rows)	0.082418	0.0849970	0.0901913	0.090001	0.092821	0.136203
3	rowCounts(X[cols, rows])	0.123533	0.1299695	0.1377765	0.137303	0.140325	0.278118

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.503183	1.472740	1.472669	1.471783	1.488112	1.289960
3	rowCounts(X[cols, rows])	2.253059	2.251977	2.249655	2.245311	2.249699	2.634017

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
1	colCounts_X_S	49.440	52.4380	53.90349	53.940	55.366	67.887
2	rowCounts_X_S	54.829	57.7135	61.24341	61.151	62.375	105.587

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts_X_S	1.109001	1.100604	1.136168	1.133686	1.126594	1.555335

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(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 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
1	colCounts_X_S	0.048826	0.049702	0.0517852	0.0518185	0.0527520	0.065642
2	colCounts(X, rows, cols)	0.075753	0.076358	0.0794161	0.0784035	0.0814750	0.096335
3	colCounts(X[rows, cols])	0.117300	0.120151	0.1262839	0.1243840	0.1275025	0.251628

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCounts(X, rows, cols)	1.551489	1.536316	1.533566	1.513041	1.544491	1.467582
3	colCounts(X[rows, cols])	2.402409	2.417428	2.438607	2.400378	2.417017	3.833338

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
1	rowCounts_X_S	0.054137	0.0551325	0.0577975	0.0587340	0.0590865	0.069548
2	rowCounts(X, cols, rows)	0.084963	0.0869950	0.0912077	0.0924395	0.0927485	0.127833
3	rowCounts(X[cols, rows])	0.127433	0.1305020	0.1386157	0.1375325	0.1398425	0.263158

	expr	min	lq	mean	median	uq	max
1	rowCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts(X, cols, rows)	1.569407	1.577926	1.578057	1.573867	1.569707	1.838054
3	rowCounts(X[cols, rows])	2.353898	2.367061	2.398300	2.341616	2.366742	3.783833

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
1	colCounts_X_S	48.826	49.7020	51.78525	51.8185	52.7520	65.642
2	rowCounts_X_S	54.137	55.1325	57.79748	58.7340	59.0865	69.548

	expr	min	lq	mean	median	uq	max
1	colCounts_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCounts_X_S	1.108774	1.109261	1.116099	1.133456	1.120081	1.059505

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 3.6.1 Patched (2019-08-27 r77078)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.3 LTS

Matrix products: default
BLAS:   /home/hb/software/R-devel/R-3-6-branch/lib/R/lib/libRblas.so
LAPACK: /home/hb/software/R-devel/R-3-6-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_1.4-6    matrixStats_0.55.0-9000 ggplot2_3.2.1          
[4] knitr_1.24              R.devices_2.16.0        R.utils_2.9.0          
[7] R.oo_1.22.0             R.methodsS3_1.7.1       history_0.0.0-9002     

loaded via a namespace (and not attached):
 [1] Biobase_2.45.0       bit64_0.9-7          splines_3.6.1       
 [4] network_1.15         assertthat_0.2.1     highr_0.8           
 [7] stats4_3.6.1         blob_1.2.0           robustbase_0.93-5   
[10] pillar_1.4.2         RSQLite_2.1.2        backports_1.1.4     
[13] lattice_0.20-38      glue_1.3.1           digest_0.6.20       
[16] colorspace_1.4-1     sandwich_2.5-1       Matrix_1.2-17       
[19] XML_3.98-1.20        lpSolve_5.6.13.3     pkgconfig_2.0.2     
[22] genefilter_1.66.0    purrr_0.3.2          ergm_3.10.4         
[25] xtable_1.8-4         mvtnorm_1.0-11       scales_1.0.0        
[28] tibble_2.1.3         annotate_1.62.0      IRanges_2.18.2      
[31] TH.data_1.0-10       withr_2.1.2          BiocGenerics_0.30.0 
[34] lazyeval_0.2.2       mime_0.7             survival_2.44-1.1   
[37] magrittr_1.5         crayon_1.3.4         statnet.common_4.3.0
[40] memoise_1.1.0        laeken_0.5.0         R.cache_0.13.0      
[43] MASS_7.3-51.4        R.rsp_0.43.1         tools_3.6.1         
[46] multcomp_1.4-10      S4Vectors_0.22.1     trust_0.1-7         
[49] munsell_0.5.0        AnnotationDbi_1.46.1 compiler_3.6.1      
[52] rlang_0.4.0          grid_3.6.1           RCurl_1.95-4.12     
[55] cwhmisc_6.6          rappdirs_0.3.1       labeling_0.3        
[58] bitops_1.0-6         base64enc_0.1-3      boot_1.3-23         
[61] gtable_0.3.0         codetools_0.2-16     DBI_1.0.0           
[64] markdown_1.1         R6_2.4.0             zoo_1.8-6           
[67] dplyr_0.8.3          bit_1.1-14           zeallot_0.1.0       
[70] parallel_3.6.1       Rcpp_1.0.2           vctrs_0.2.0         
[73] DEoptimR_1.0-8       tidyselect_0.2.5     xfun_0.9            
[76] coda_0.19-3         

Total processing time was 33.66 secs.

Reproducibility

To reproduce this report, do:

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

---

Copyright Dongcan Jiang. Last updated on 2019-09-10 20:36:47 (-0700 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>