matrixStats.benchmarks

---

colCummins() and rowCummins() benchmarks on subsetted computation

This report benchmark the performance of colCummins() and rowCummins() on subsetted computation.

Data type “integer”

Data

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

Results

10x10 integer matrix

> X <- data[["10x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5193783 277.4    8529671 455.6  8529671 455.6
Vcells 9739225  74.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5185420 277.0    8529671 455.6  8529671 455.6
Vcells 9711945  74.1   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.001723	0.0019105	0.0029729	0.0019780	0.0020995	0.097800
2	colCummins(X, rows, cols)	0.001967	0.0022870	0.0024131	0.0023620	0.0024365	0.004847
3	colCummins(X[rows, cols])	0.002574	0.0029750	0.0031610	0.0030825	0.0032120	0.010422

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colCummins(X, rows, cols)	1.141614	1.197069	0.811703	1.194136	1.160514	0.0495603
3	colCummins(X[rows, cols])	1.493906	1.557184	1.063268	1.558392	1.529888	0.1065644

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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	rowCummins_X_S	0.001747	0.0020055	0.0021237	0.0020825	0.0021585	0.004422
2	rowCummins(X, cols, rows)	0.002125	0.0023695	0.0033469	0.0024725	0.0025790	0.085197
3	rowCummins(X[cols, rows])	0.002817	0.0030770	0.0032745	0.0031445	0.0032790	0.009464

	expr	min	lq	mean	median	uq	max
1	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins(X, cols, rows)	1.216371	1.181501	1.575995	1.187275	1.194811	19.266621
3	rowCummins(X[cols, rows])	1.612478	1.534281	1.541918	1.509964	1.519111	2.140208

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

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

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.723	1.9105	2.97291	1.9780	2.0995	97.800
2	rowCummins_X_S	1.747	2.0055	2.12368	2.0825	2.1585	4.422

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	rowCummins_X_S	1.013929	1.049725	0.7143439	1.052831	1.028102	0.0452147

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

100x100 integer matrix

> X <- data[["100x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5184048 276.9    8529671 455.6  8529671 455.6
Vcells 9380529  71.6   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5184024 276.9    8529671 455.6  8529671 455.6
Vcells 9385582  71.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.015989	0.016298	0.0165247	0.0163960	0.0165695	0.021165
2	colCummins(X, rows, cols)	0.016327	0.016922	0.0172117	0.0170985	0.0172670	0.021670
3	colCummins(X[rows, cols])	0.027582	0.027799	0.0282629	0.0278985	0.0280635	0.052871

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCummins(X, rows, cols)	1.021140	1.038287	1.041576	1.042846	1.042095	1.023860
3	colCummins(X[rows, cols])	1.725061	1.705669	1.710340	1.701543	1.693684	2.498039

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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	rowCummins_X_S	0.020410	0.0214990	0.0222479	0.0223045	0.0224610	0.036251
2	rowCummins(X, cols, rows)	0.022831	0.0237855	0.0247767	0.0244665	0.0246990	0.052317
3	rowCummins(X[cols, rows])	0.030901	0.0325430	0.0333857	0.0336010	0.0338715	0.038851

	expr	min	lq	mean	median	uq	max
1	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins(X, cols, rows)	1.118618	1.106354	1.113663	1.096931	1.099639	1.443188
3	rowCummins(X[cols, rows])	1.514013	1.513698	1.500623	1.506467	1.508014	1.071722

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

Table: Benchmarking of colCummins_X_S() and rowCummins_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	colCummins_X_S	15.989	16.298	16.52470	16.3960	16.5695	21.165
2	rowCummins_X_S	20.410	21.499	22.24791	22.3045	22.4610	36.251

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins_X_S	1.276503	1.319119	1.346343	1.360362	1.355563	1.712781

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

1000x10 integer matrix

> X <- data[["1000x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5184790 276.9    8529671 455.6  8529671 455.6
Vcells 9384580  71.6   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5184766 276.9    8529671 455.6  8529671 455.6
Vcells 9389633  71.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.015677	0.0160335	0.0162090	0.0161375	0.0162595	0.017855
2	colCummins(X, rows, cols)	0.017920	0.0184930	0.0187301	0.0186485	0.0189205	0.023008
3	colCummins(X[rows, cols])	0.027705	0.0281665	0.0287674	0.0282615	0.0285090	0.054637

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCummins(X, rows, cols)	1.143076	1.153398	1.155539	1.155600	1.163658	1.288603
3	colCummins(X[rows, cols])	1.767239	1.756728	1.774782	1.751294	1.753375	3.060039

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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	rowCummins_X_S	0.020951	0.0219075	0.0227796	0.0221265	0.0233110	0.036048
2	rowCummins(X, cols, rows)	0.024800	0.0259840	0.0272776	0.0269895	0.0280640	0.042307
3	rowCummins(X[cols, rows])	0.033155	0.0347335	0.0361015	0.0355385	0.0370445	0.070689

	expr	min	lq	mean	median	uq	max
1	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins(X, cols, rows)	1.183714	1.186078	1.197455	1.219782	1.203895	1.173630
3	rowCummins(X[cols, rows])	1.582502	1.585462	1.584816	1.606151	1.589142	1.960969

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

Table: Benchmarking of colCummins_X_S() and rowCummins_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	colCummins_X_S	15.677	16.0335	16.20898	16.1375	16.2595	17.855
2	rowCummins_X_S	20.951	21.9075	22.77964	22.1265	23.3110	36.048

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
2	rowCummins_X_S	1.336416	1.366358	1.405372	1.371123	1.433685	2.01893

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

10x1000 integer matrix

> X <- data[["10x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5184995 277.0    8529671 455.6  8529671 455.6
Vcells 9385410  71.7   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5184971 277.0    8529671 455.6  8529671 455.6
Vcells 9390463  71.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.015699	0.0165555	0.0173917	0.0171700	0.0177065	0.032679
2	colCummins(X, rows, cols)	0.018502	0.0190995	0.0199065	0.0196435	0.0203165	0.036713
3	colCummins(X[rows, cols])	0.028325	0.0298230	0.0311322	0.0310330	0.0315580	0.046307

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCummins(X, rows, cols)	1.178546	1.153665	1.144600	1.144059	1.147404	1.123443
3	colCummins(X[rows, cols])	1.804255	1.801395	1.790059	1.807397	1.782283	1.417026

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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	rowCummins_X_S	0.018412	0.0193380	0.0197912	0.0199035	0.0202200	0.023383
2	rowCummins(X, cols, rows)	0.022895	0.0238335	0.0249646	0.0245585	0.0251085	0.055176
3	rowCummins(X[cols, rows])	0.029324	0.0306865	0.0316800	0.0319300	0.0323295	0.045082

	expr	min	lq	mean	median	uq	max
1	rowCummins_X_S	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
2	rowCummins(X, cols, rows)	1.243483	1.23247	1.261397	1.233879	1.241766	2.359663
3	rowCummins(X[cols, rows])	1.592657	1.58685	1.600711	1.604240	1.598887	1.927982

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

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

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	15.699	16.5555	17.39171	17.1700	17.7065	32.679
2	rowCummins_X_S	18.412	19.3380	19.79123	19.9035	20.2200	23.383

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins_X_S	1.172814	1.168071	1.137969	1.159202	1.141954	0.715536

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

100x1000 integer matrix

> X <- data[["100x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5185206 277.0    8529671 455.6  8529671 455.6
Vcells 9408082  71.8   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5185182 277.0    8529671 455.6  8529671 455.6
Vcells 9458135  72.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.098381	0.1102525	0.1204229	0.1144985	0.1323645	0.254618
2	colCummins(X, rows, cols)	0.100008	0.1119840	0.1238642	0.1165325	0.1336210	0.310245
3	colCummins(X[rows, cols])	0.171976	0.1934190	0.2130575	0.2027355	0.2365120	0.317682

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCummins(X, rows, cols)	1.016538	1.015705	1.028577	1.017764	1.009493	1.218472
3	colCummins(X[rows, cols])	1.748061	1.754328	1.769245	1.770639	1.786823	1.247681

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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	rowCummins_X_S	0.115840	0.1361965	0.1505212	0.1457645	0.163909	0.206026
2	rowCummins(X, cols, rows)	0.130276	0.1420105	0.1625764	0.1589730	0.172125	0.262708
3	rowCummins(X[cols, rows])	0.178784	0.2015465	0.2292682	0.2246175	0.248731	0.308502

	expr	min	lq	mean	median	uq	max
1	rowCummins_X_S	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins(X, cols, rows)	1.12462	1.042688	1.080090	1.090615	1.050125	1.275121
3	rowCummins(X[cols, rows])	1.54337	1.479821	1.523162	1.540962	1.517494	1.497394

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

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

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	98.381	110.2525	120.4229	114.4985	132.3645	254.618
2	rowCummins_X_S	115.840	136.1965	150.5212	145.7645	163.9090	206.026

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCummins_X_S	1.177463	1.235314	1.249939	1.273069	1.238315	0.8091572

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

1000x100 integer matrix

> X <- data[["1000x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5185416 277.0    8529671 455.6  8529671 455.6
Vcells 9408858  71.8   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5185392 277.0    8529671 455.6  8529671 455.6
Vcells 9458911  72.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.096354	0.1056810	0.1136942	0.1096075	0.1164960	0.160298
2	colCummins(X, rows, cols)	0.097761	0.1065750	0.1173395	0.1132075	0.1229335	0.169410
3	colCummins(X[rows, cols])	0.166799	0.1833995	0.2014776	0.1953445	0.2171025	0.306536

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCummins(X, rows, cols)	1.014602	1.008459	1.032063	1.032844	1.055259	1.056844
3	colCummins(X[rows, cols])	1.731106	1.735407	1.772101	1.782218	1.863605	1.912288

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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	rowCummins_X_S	0.123129	0.1338415	0.1548275	0.1513005	0.1657165	0.211283
2	rowCummins(X, cols, rows)	0.131819	0.1511770	0.1683311	0.1656555	0.1793525	0.282531
3	rowCummins(X[cols, rows])	0.189457	0.2080270	0.2382065	0.2308440	0.2620870	0.322832

	expr	min	lq	mean	median	uq	max
1	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins(X, cols, rows)	1.070576	1.129523	1.087217	1.094877	1.082285	1.337216
3	rowCummins(X[cols, rows])	1.538687	1.554279	1.538529	1.525732	1.581538	1.527960

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

Table: Benchmarking of colCummins_X_S() and rowCummins_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	colCummins_X_S	96.354	105.6810	113.6942	109.6075	116.4960	160.298
2	rowCummins_X_S	123.129	133.8415	154.8275	151.3005	165.7165	211.283

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins_X_S	1.277882	1.266467	1.361788	1.380385	1.422508	1.318064

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

Data type “double”

Data

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

Results

10x10 double matrix

> X <- data[["10x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5185634 277.0    8529671 455.6  8529671 455.6
Vcells 9499964  72.5   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5185601 277.0    8529671 455.6  8529671 455.6
Vcells 9500102  72.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.001894	0.0020290	0.0022981	0.0020830	0.0021460	0.018771
2	colCummins(X, rows, cols)	0.002096	0.0024145	0.0026213	0.0024590	0.0025495	0.011467
3	colCummins(X[rows, cols])	0.002890	0.0030955	0.0033837	0.0032215	0.0033135	0.013518

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colCummins(X, rows, cols)	1.106653	1.189995	1.140662	1.180509	1.188024	0.6108891
3	colCummins(X[rows, cols])	1.525871	1.525628	1.472420	1.546568	1.544035	0.7201534

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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	rowCummins_X_S	0.001817	0.0020125	0.0021330	0.0020940	0.0021655	0.004287
2	rowCummins(X, cols, rows)	0.002170	0.0023525	0.0026411	0.0024175	0.0025500	0.019917
3	rowCummins(X[cols, rows])	0.002758	0.0030680	0.0032030	0.0031530	0.0033130	0.005159

	expr	min	lq	mean	median	uq	max
1	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins(X, cols, rows)	1.194276	1.168944	1.238203	1.154489	1.177557	4.645906
3	rowCummins(X[cols, rows])	1.517887	1.524472	1.501653	1.505731	1.529901	1.203406

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

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

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.894	2.0290	2.29806	2.083	2.1460	18.771
2	rowCummins_X_S	1.817	2.0125	2.13301	2.094	2.1655	4.287

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.0000000	1.0000000	1.0000000	1.000000	1.000000	1.0000000
2	rowCummins_X_S	0.9593453	0.9918679	0.9281786	1.005281	1.009087	0.2283842

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

100x100 double matrix

> X <- data[["100x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5185833 277.0    8529671 455.6  8529671 455.6
Vcells 9505916  72.6   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5185809 277.0    8529671 455.6  8529671 455.6
Vcells 9515969  72.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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
2	colCummins(X, rows, cols)	0.030582	0.0321045	0.0329631	0.0326850	0.0331580	0.047132
1	colCummins_X_S	0.030731	0.0318760	0.0330078	0.0330270	0.0331895	0.048927
3	colCummins(X[rows, cols])	0.046103	0.0490820	0.0495816	0.0494065	0.0496925	0.074218

	expr	min	lq	mean	median	uq	max
2	colCummins(X, rows, cols)	1.000000	1.0000000	1.000000	1.000000	1.000000	1.000000
1	colCummins_X_S	1.004872	0.9928826	1.001354	1.010463	1.000950	1.038085
3	colCummins(X[rows, cols])	1.507521	1.5288199	1.504154	1.511596	1.498658	1.574684

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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
2	rowCummins(X, cols, rows)	0.018296	0.0187390	0.0201161	0.0189495	0.0196400	0.065204
1	rowCummins_X_S	0.019914	0.0202275	0.0212950	0.0204705	0.0214050	0.031559
3	rowCummins(X[cols, rows])	0.036230	0.0365095	0.0384711	0.0367665	0.0377755	0.072528

	expr	min	lq	mean	median	uq	max
2	rowCummins(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	rowCummins_X_S	1.088435	1.079433	1.058605	1.080266	1.089868	0.484004
3	rowCummins(X[cols, rows])	1.980214	1.948316	1.912458	1.940236	1.923396	1.112324

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

Table: Benchmarking of colCummins_X_S() and rowCummins_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	rowCummins_X_S	19.914	20.2275	21.29496	20.4705	21.4050	31.559
1	colCummins_X_S	30.731	31.8760	33.00777	33.0270	33.1895	48.927

	expr	min	lq	mean	median	uq	max
2	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCummins_X_S	1.543186	1.575874	1.550027	1.613395	1.550549	1.550334

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

1000x10 double matrix

> X <- data[["1000x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5186032 277.0    8529671 455.6  8529671 455.6
Vcells 9507323  72.6   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5186008 277.0    8529671 455.6  8529671 455.6
Vcells 9517376  72.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.027322	0.0286270	0.0296818	0.0296685	0.0298270	0.043600
2	colCummins(X, rows, cols)	0.028195	0.0293805	0.0304062	0.0307565	0.0310990	0.036943
3	colCummins(X[rows, cols])	0.042789	0.0445085	0.0458371	0.0466285	0.0468955	0.072171

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colCummins(X, rows, cols)	1.031952	1.026321	1.024405	1.036672	1.042646	0.8473165
3	colCummins(X[rows, cols])	1.566101	1.554774	1.544283	1.571650	1.572250	1.6552982

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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	rowCummins_X_S	0.020104	0.0211380	0.0214802	0.0214425	0.021652	0.026957
2	rowCummins(X, cols, rows)	0.020889	0.0220675	0.0226812	0.0225090	0.022817	0.036190
3	rowCummins(X[cols, rows])	0.038867	0.0400180	0.0416119	0.0413045	0.041519	0.077111

	expr	min	lq	mean	median	uq	max
1	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins(X, cols, rows)	1.039047	1.043973	1.055914	1.049738	1.053806	1.342508
3	rowCummins(X[cols, rows])	1.933297	1.893178	1.937225	1.926291	1.917560	2.860519

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

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

	expr	min	lq	mean	median	uq	max
2	rowCummins_X_S	20.104	21.138	21.48018	21.4425	21.652	26.957
1	colCummins_X_S	27.322	28.627	29.68183	29.6685	29.827	43.600

	expr	min	lq	mean	median	uq	max
2	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCummins_X_S	1.359033	1.354291	1.381824	1.383631	1.377563	1.617391

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

10x1000 double matrix

> X <- data[["10x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5186237 277.0    8529671 455.6  8529671 455.6
Vcells 9507459  72.6   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5186213 277.0    8529671 455.6  8529671 455.6
Vcells 9517512  72.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.036557	0.0395040	0.0414634	0.0414630	0.042958	0.062859
2	colCummins(X, rows, cols)	0.039742	0.0434285	0.0470496	0.0466835	0.049544	0.076442
3	colCummins(X[rows, cols])	0.055167	0.0590315	0.0610964	0.0615955	0.062958	0.078570

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCummins(X, rows, cols)	1.087124	1.099344	1.134726	1.125907	1.153312	1.216087
3	colCummins(X[rows, cols])	1.509068	1.494317	1.473502	1.485553	1.465571	1.249940

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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	rowCummins_X_S	0.017213	0.0182420	0.0186094	0.018473	0.0186205	0.032489
2	rowCummins(X, cols, rows)	0.018850	0.0199065	0.0208403	0.020239	0.0205645	0.049414
3	rowCummins(X[cols, rows])	0.033322	0.0347830	0.0355478	0.035540	0.0357255	0.044009

	expr	min	lq	mean	median	uq	max
1	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCummins(X, cols, rows)	1.095103	1.091246	1.119880	1.095599	1.104401	1.520946
3	rowCummins(X[cols, rows])	1.935862	1.906754	1.910211	1.923889	1.918611	1.354582

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

Table: Benchmarking of colCummins_X_S() and rowCummins_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	rowCummins_X_S	17.213	18.242	18.60938	18.473	18.6205	32.489
1	colCummins_X_S	36.557	39.504	41.46342	41.463	42.9580	62.859

	expr	min	lq	mean	median	uq	max
2	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCummins_X_S	2.123802	2.165552	2.228092	2.244519	2.307027	1.934778

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

100x1000 double matrix

> X <- data[["100x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5186448 277.0    8529671 455.6  8529671 455.6
Vcells 9552913  72.9   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5186424 277.0    8529671 455.6  8529671 455.6
Vcells 9652966  73.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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	colCummins_X_S	0.192922	0.1942375	0.2320334	0.2137830	0.2517620	0.338854
2	colCummins(X, rows, cols)	0.189785	0.1913770	0.2293150	0.2188695	0.2507695	0.409263
3	colCummins(X[rows, cols])	0.287755	0.2975090	0.3521140	0.3477515	0.3781630	0.480243

	expr	min	lq	mean	median	uq	max
1	colCummins_X_S	1.0000000	1.0000000	1.0000000	1.000000	1.0000000	1.000000
2	colCummins(X, rows, cols)	0.9837395	0.9852732	0.9882845	1.023793	0.9960578	1.207786
3	colCummins(X[rows, cols])	1.4915614	1.5316764	1.5175145	1.626656	1.5020654	1.417256

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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
2	rowCummins(X, cols, rows)	0.100402	0.1087440	0.1224488	0.1178120	0.1288175	0.277172
1	rowCummins_X_S	0.109549	0.1238685	0.1382438	0.1340865	0.1492380	0.206046
3	rowCummins(X[cols, rows])	0.209804	0.2298560	0.2583808	0.2498130	0.2828965	0.377144

	expr	min	lq	mean	median	uq	max
2	rowCummins(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowCummins_X_S	1.091104	1.139083	1.128993	1.138140	1.158523	0.7433868
3	rowCummins(X[cols, rows])	2.089640	2.113735	2.110114	2.120438	2.196103	1.3606858

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

Table: Benchmarking of colCummins_X_S() and rowCummins_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
2	rowCummins_X_S	109.549	123.8685	138.2438	134.0865	149.238	206.046
1	colCummins_X_S	192.922	194.2375	232.0334	213.7830	251.762	338.854

	expr	min	lq	mean	median	uq	max
2	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCummins_X_S	1.761057	1.568094	1.678435	1.594366	1.686983	1.644555

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

1000x100 double matrix

> X <- data[["1000x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5186658 277.0    8529671 455.6  8529671 455.6
Vcells 9553054  72.9   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCummins_X_S = colCummins(X_S), `colCummins(X, rows, cols)` = colCummins(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCummins(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5186634 277.0    8529671 455.6  8529671 455.6
Vcells 9653107  73.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCummins_X_S = rowCummins(X_S), `rowCummins(X, cols, rows)` = rowCummins(X, 
+     rows = cols, cols = rows), `rowCummins(X[cols, rows])` = rowCummins(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCummins_X_S(), colCummins(X, rows, cols)() and colCummins(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
2	colCummins(X, rows, cols)	0.160836	0.1699880	0.1981516	0.1933905	0.2113140	0.332304
1	colCummins_X_S	0.169699	0.1742625	0.2035654	0.1952975	0.2206660	0.299149
3	colCummins(X[rows, cols])	0.262596	0.2725995	0.3240394	0.3233140	0.3543425	0.536329

	expr	min	lq	mean	median	uq	max
2	colCummins(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colCummins_X_S	1.055106	1.025146	1.027321	1.009861	1.044256	0.9002269
3	colCummins(X[rows, cols])	1.632694	1.603640	1.635310	1.671820	1.676853	1.6139709

Table: Benchmarking of rowCummins_X_S(), rowCummins(X, cols, rows)() and rowCummins(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
2	rowCummins(X, cols, rows)	0.102477	0.115312	0.1314510	0.126565	0.1409870	0.289618
1	rowCummins_X_S	0.115118	0.122562	0.1427030	0.138125	0.1556825	0.206438
3	rowCummins(X[cols, rows])	0.216508	0.232242	0.2684032	0.257272	0.2953305	0.456358

	expr	min	lq	mean	median	uq	max
2	rowCummins(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowCummins_X_S	1.123355	1.062873	1.085598	1.091336	1.104233	0.7127941
3	rowCummins(X[cols, rows])	2.112747	2.014032	2.041850	2.032726	2.094736	1.5757239

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

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

	expr	min	lq	mean	median	uq	max
2	rowCummins_X_S	115.118	122.5620	142.7030	138.1250	155.6825	206.438
1	colCummins_X_S	169.699	174.2625	203.5654	195.2975	220.6660	299.149

	expr	min	lq	mean	median	uq	max
2	rowCummins_X_S	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000
1	colCummins_X_S	1.474131	1.421831	1.426497	1.413919	1.41741	1.449099

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

Appendix

Session information

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

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

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

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

other attached packages:
[1] microbenchmark_1.4-7   matrixStats_0.60.1     ggplot2_3.3.5         
[4] knitr_1.33             R.devices_2.17.0       R.utils_2.10.1        
[7] R.oo_1.24.0            R.methodsS3_1.8.1-9001 history_0.0.1-9000    

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

Total processing time was 22.77 secs.

Reproducibility

To reproduce this report, do:

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

---

Copyright Dongcan Jiang. Last updated on 2021-08-25 18:52:42 (+0200 UTC). Powered by RSP.