matrixStats.benchmarks

---

colSums2() and rowSums2() benchmarks on subsetted computation

This report benchmark the performance of colSums2() and rowSums2() 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  5297702 283.0    8529671 455.6  8529671 455.6
Vcells 10459475  79.8   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5289561 282.5    8529671 455.6  8529671 455.6
Vcells 10432899  79.6   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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	colSums2_X_S	0.001917	0.0019960	0.0029790	0.0020690	0.0021355	0.090851
2	colSums2(X, rows, cols)	0.002327	0.0024455	0.0025583	0.0025005	0.0026000	0.004799
3	colSums2(X[rows, cols])	0.002805	0.0029560	0.0031580	0.0030515	0.0031725	0.007481

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	colSums2(X, rows, cols)	1.213876	1.225200	0.8587637	1.208555	1.217513	0.0528228
3	colSums2(X[rows, cols])	1.463224	1.480962	1.0600762	1.474867	1.485601	0.0823436

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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	rowSums2_X_S	0.001868	0.0019700	0.0020646	0.0020035	0.0021170	0.004339
2	rowSums2(X, cols, rows)	0.002282	0.0023745	0.0033643	0.0024320	0.0025395	0.091039
3	rowSums2(X[cols, rows])	0.002685	0.0029045	0.0031199	0.0030495	0.0031595	0.005453

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.221627	1.205330	1.629514	1.213876	1.199575	20.981563
3	rowSums2(X[cols, rows])	1.437366	1.474365	1.511133	1.522086	1.492442	1.256741

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+10x10 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_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	rowSums2_X_S	1.868	1.970	2.06461	2.0035	2.1170	4.339
1	colSums2_X_S	1.917	1.996	2.97905	2.0690	2.1355	90.851

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
1	colSums2_X_S	1.026231	1.013198	1.442912	1.032693	1.008739	20.93823

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5288183 282.5    8529671 455.6  8529671 455.6
Vcells 10101777  77.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5288159 282.5    8529671 455.6  8529671 455.6
Vcells 10106830  77.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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
2	colSums2(X, rows, cols)	0.015154	0.0158085	0.0162980	0.0160540	0.0163710	0.030804
1	colSums2_X_S	0.016126	0.0164790	0.0170535	0.0170285	0.0172215	0.021129
3	colSums2(X[rows, cols])	0.026489	0.0274090	0.0286065	0.0279745	0.0289585	0.057307

	expr	min	lq	mean	median	uq	max
2	colSums2(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colSums2_X_S	1.064142	1.042414	1.046356	1.060701	1.051952	0.6859174
3	colSums2(X[rows, cols])	1.747987	1.733814	1.755221	1.742525	1.768890	1.8603753

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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
2	rowSums2(X, cols, rows)	0.015200	0.0159045	0.0167424	0.0161240	0.0166115	0.037784
1	rowSums2_X_S	0.015452	0.0162115	0.0166356	0.0164465	0.0169220	0.022989
3	rowSums2(X[cols, rows])	0.025896	0.0271610	0.0279388	0.0274740	0.0284070	0.036413

	expr	min	lq	mean	median	uq	max
2	rowSums2(X, cols, rows)	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
1	rowSums2_X_S	1.016579	1.019303	0.9936192	1.020001	1.018692	0.6084321
3	rowSums2(X[cols, rows])	1.703684	1.707756	1.6687390	1.703920	1.710080	0.9637148

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+100x100 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+100x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	15.452	16.2115	16.63562	16.4465	16.9220	22.989
1	colSums2_X_S	16.126	16.4790	17.05348	17.0285	17.2215	21.129

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colSums2_X_S	1.043619	1.016501	1.025118	1.035387	1.017699	0.9190917

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5288926 282.5    8529671 455.6  8529671 455.6
Vcells 10105817  77.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5288902 282.5    8529671 455.6  8529671 455.6
Vcells 10110870  77.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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	colSums2_X_S	0.015116	0.0157575	0.0160127	0.0159035	0.0163930	0.019119
2	colSums2(X, rows, cols)	0.016825	0.0177845	0.0185554	0.0181820	0.0186565	0.031995
3	colSums2(X[rows, cols])	0.026027	0.0266035	0.0280549	0.0273760	0.0283605	0.057529

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.113059	1.128637	1.158789	1.143270	1.138077	1.673466
3	colSums2(X[rows, cols])	1.721818	1.688307	1.752043	1.721382	1.730037	3.008996

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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	rowSums2_X_S	0.015052	0.0156505	0.0159805	0.0157925	0.0163245	0.022156
2	rowSums2(X, cols, rows)	0.016871	0.0177090	0.0184046	0.0180600	0.0186610	0.033190
3	rowSums2(X[cols, rows])	0.027499	0.0285005	0.0301481	0.0292280	0.0304665	0.062600

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.120848	1.131529	1.151692	1.143581	1.143128	1.498014
3	rowSums2(X[cols, rows])	1.826933	1.821060	1.886553	1.850752	1.866305	2.825420

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+1000x10 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on integer+1000x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	15.052	15.6505	15.98053	15.7925	16.3245	22.156
1	colSums2_X_S	15.116	15.7575	16.01271	15.9035	16.3930	19.119

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colSums2_X_S	1.004252	1.006837	1.002014	1.007029	1.004196	0.8629265

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5289131 282.5    8529671 455.6  8529671 455.6
Vcells 10106668  77.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5289107 282.5    8529671 455.6  8529671 455.6
Vcells 10111721  77.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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	colSums2_X_S	0.018567	0.0193270	0.0200568	0.0201125	0.0203765	0.037350
2	colSums2(X, rows, cols)	0.019119	0.0201590	0.0211767	0.0208620	0.0213490	0.035923
3	colSums2(X[rows, cols])	0.031058	0.0328255	0.0337760	0.0334745	0.0344865	0.043080

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colSums2(X, rows, cols)	1.029730	1.043049	1.055837	1.037265	1.047727	0.9617938
3	colSums2(X[rows, cols])	1.672753	1.698427	1.684017	1.664363	1.692464	1.1534137

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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	rowSums2_X_S	0.018112	0.0188490	0.0196013	0.0194570	0.020124	0.028014
2	rowSums2(X, cols, rows)	0.018698	0.0194175	0.0204472	0.0198375	0.020475	0.043371
3	rowSums2(X[cols, rows])	0.029081	0.0302220	0.0316607	0.0312250	0.032506	0.045241

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.032354	1.030161	1.043157	1.019556	1.017442	1.548190
3	rowSums2(X[cols, rows])	1.605621	1.603374	1.615237	1.604821	1.615285	1.614942

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+10x1000 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_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	rowSums2_X_S	18.112	18.849	19.60130	19.4570	20.1240	28.014
1	colSums2_X_S	18.567	19.327	20.05681	20.1125	20.3765	37.350

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000
1	colSums2_X_S	1.025122	1.025359	1.023239	1.03369	1.012547	1.333262

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5289341 282.5    8529671 455.6  8529671 455.6
Vcells 10129340  77.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5289317 282.5    8529671 455.6  8529671 455.6
Vcells 10179393  77.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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
2	colSums2(X, rows, cols)	0.087365	0.095771	0.1067721	0.1043025	0.1149665	0.191974
1	colSums2_X_S	0.098938	0.111564	0.1245812	0.1217165	0.1343105	0.183901
3	colSums2(X[rows, cols])	0.167719	0.192112	0.2119592	0.2058850	0.2274205	0.323292

	expr	min	lq	mean	median	uq	max
2	colSums2(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colSums2_X_S	1.132467	1.164904	1.166796	1.166957	1.168258	0.9579474
3	colSums2(X[rows, cols])	1.919750	2.005952	1.985155	1.973922	1.978146	1.6840405

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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
2	rowSums2(X, cols, rows)	0.093528	0.0991485	0.1119656	0.1089325	0.1198325	0.201175
1	rowSums2_X_S	0.097333	0.1091325	0.1213618	0.1192595	0.1308295	0.199279
3	rowSums2(X[cols, rows])	0.167777	0.1828180	0.2023512	0.1965940	0.2187180	0.291069

	expr	min	lq	mean	median	uq	max
2	rowSums2(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowSums2_X_S	1.040683	1.100697	1.083921	1.094802	1.091770	0.9905754
3	rowSums2(X[cols, rows])	1.793869	1.843881	1.807263	1.804732	1.825198	1.4468448

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+100x1000 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_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	rowSums2_X_S	97.333	109.1325	121.3618	119.2595	130.8295	199.279
1	colSums2_X_S	98.938	111.5640	124.5812	121.7165	134.3105	183.901

	expr	min	lq	mean	median	uq	max
2	rowSums2_X_S	1.00000	1.00000	1.000000	1.000000	1.000000	1.0000000
1	colSums2_X_S	1.01649	1.02228	1.026528	1.020602	1.026607	0.9228318

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5289554 282.5    8529671 455.6  8529671 455.6
Vcells 10130119  77.3   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5289530 282.5    8529671 455.6  8529671 455.6
Vcells 10180172  77.7   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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
2	colSums2(X, rows, cols)	0.089190	0.0971085	0.1077117	0.1063650	0.1151745	0.147362
1	colSums2_X_S	0.098186	0.1071025	0.1158809	0.1140255	0.1221325	0.165320
3	colSums2(X[rows, cols])	0.166742	0.1779460	0.2024797	0.1974365	0.2228625	0.305317

	expr	min	lq	mean	median	uq	max
2	colSums2(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colSums2_X_S	1.100863	1.102916	1.075844	1.072021	1.060413	1.121863
3	colSums2(X[rows, cols])	1.869514	1.832445	1.879830	1.856217	1.934999	2.071884

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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
2	rowSums2(X, cols, rows)	0.094610	0.106023	0.1165371	0.1124645	0.1242875	0.214450
1	rowSums2_X_S	0.101942	0.110439	0.1234676	0.1217585	0.1329315	0.170933
3	rowSums2(X[cols, rows])	0.170344	0.183902	0.2053379	0.2025215	0.2211970	0.288600

	expr	min	lq	mean	median	uq	max
2	rowSums2(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowSums2_X_S	1.077497	1.041651	1.059470	1.082639	1.069548	0.7970762
3	rowSums2(X[cols, rows])	1.800486	1.734548	1.761996	1.800759	1.779720	1.3457682

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on integer+1000x100 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_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	colSums2_X_S	98.186	107.1025	115.8809	114.0255	122.1325	165.320
2	rowSums2_X_S	101.942	110.4390	123.4677	121.7585	132.9315	170.933

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.00000	1.000000	1.00000	1.000000
2	rowSums2_X_S	1.038254	1.031152	1.06547	1.067818	1.08842	1.033952

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5289771 282.6    8529671 455.6  8529671 455.6
Vcells 10221223  78.0   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5289738 282.6    8529671 455.6  8529671 455.6
Vcells 10221361  78.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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	colSums2_X_S	0.001904	0.0019810	0.0022126	0.002029	0.0021055	0.017383
2	colSums2(X, rows, cols)	0.002305	0.0023955	0.0025080	0.002454	0.0025215	0.004870
3	colSums2(X[rows, cols])	0.002866	0.0030730	0.0032465	0.003152	0.0033060	0.007907

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colSums2(X, rows, cols)	1.210609	1.209238	1.133485	1.209463	1.197578	0.2801588
3	colSums2(X[rows, cols])	1.505252	1.551237	1.467254	1.553475	1.570173	0.4548697

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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	rowSums2_X_S	0.001906	0.0020000	0.0021061	0.0020445	0.002148	0.004407
2	rowSums2(X, cols, rows)	0.002337	0.0024335	0.0027397	0.0024790	0.002594	0.022945
3	rowSums2(X[cols, rows])	0.002836	0.0031030	0.0032540	0.0031650	0.003266	0.007526

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.226128	1.21675	1.300821	1.212521	1.207635	5.206490
3	rowSums2(X[cols, rows])	1.487933	1.55150	1.545050	1.548056	1.520484	1.707738

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+10x10 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_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	colSums2_X_S	1.904	1.981	2.21261	2.0290	2.1055	17.383
2	rowSums2_X_S	1.906	2.000	2.10610	2.0445	2.1480	4.407

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.00000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	rowSums2_X_S	1.00105	1.009591	0.9518623	1.007639	1.020185	0.2535236

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5289968 282.6    8529671 455.6  8529671 455.6
Vcells 10227171  78.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5289944 282.6    8529671 455.6  8529671 455.6
Vcells 10237224  78.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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	colSums2_X_S	0.013410	0.0139380	0.0146905	0.0142065	0.0145895	0.028955
2	colSums2(X, rows, cols)	0.016157	0.0168385	0.0173912	0.0172440	0.0176265	0.023075
3	colSums2(X[rows, cols])	0.028664	0.0297690	0.0311193	0.0300710	0.0310970	0.063370

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colSums2(X, rows, cols)	1.204847	1.208100	1.183840	1.213811	1.208163	0.7969263
3	colSums2(X[rows, cols])	2.137509	2.135816	2.118327	2.116707	2.131464	2.1885685

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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	rowSums2_X_S	0.014661	0.0155850	0.0164210	0.0159010	0.0163445	0.035258
2	rowSums2(X, cols, rows)	0.016131	0.0170090	0.0178754	0.0173925	0.0178840	0.044955
3	rowSums2(X[cols, rows])	0.030101	0.0314105	0.0325872	0.0316670	0.0327950	0.058305

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.100266	1.091370	1.088567	1.093799	1.094191	1.275030
3	rowSums2(X[cols, rows])	2.053134	2.015432	1.984478	1.991510	2.006485	1.653667

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+100x100 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+100x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	13.410	13.938	14.69049	14.2065	14.5895	28.955
2	rowSums2_X_S	14.661	15.585	16.42104	15.9010	16.3445	35.258

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2_X_S	1.093289	1.118166	1.117801	1.119276	1.120292	1.217683

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5290168 282.6    8529671 455.6  8529671 455.6
Vcells 10228580  78.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5290144 282.6    8529671 455.6  8529671 455.6
Vcells 10238633  78.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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	colSums2_X_S	0.011602	0.0120325	0.0123628	0.0122770	0.0126705	0.016223
2	colSums2(X, rows, cols)	0.017080	0.0177885	0.0183413	0.0181900	0.0188215	0.023480
3	colSums2(X[rows, cols])	0.027325	0.0277930	0.0294417	0.0287365	0.0299745	0.059768

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.472160	1.478371	1.483591	1.481632	1.485458	1.447328
3	colSums2(X[rows, cols])	2.355197	2.309828	2.381478	2.340678	2.365692	3.684152

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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	rowSums2_X_S	0.013388	0.0138920	0.0143209	0.014225	0.0147095	0.020348
2	rowSums2(X, cols, rows)	0.017153	0.0178305	0.0184661	0.018298	0.0188890	0.024655
3	rowSums2(X[cols, rows])	0.031939	0.0323180	0.0343460	0.033483	0.0348500	0.083655

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.281222	1.283508	1.289454	1.286327	1.284136	1.211667
3	rowSums2(X[cols, rows])	2.385644	2.326375	2.398321	2.353814	2.369217	4.111215

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+1000x10 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_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	colSums2_X_S	11.602	12.0325	12.36277	12.277	12.6705	16.223
2	rowSums2_X_S	13.388	13.8920	14.32086	14.225	14.7095	20.348

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
2	rowSums2_X_S	1.153939	1.15454	1.158386	1.158671	1.160925	1.254269

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5290373 282.6    8529671 455.6  8529671 455.6
Vcells 10228716  78.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5290349 282.6    8529671 455.6  8529671 455.6
Vcells 10238769  78.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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	colSums2_X_S	0.017197	0.0177610	0.0187146	0.0184275	0.0188005	0.039869
2	colSums2(X, rows, cols)	0.020360	0.0209175	0.0222504	0.0215050	0.0220520	0.043881
3	colSums2(X[rows, cols])	0.034517	0.0356810	0.0365928	0.0361145	0.0373210	0.056217

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.183927	1.177721	1.188936	1.167006	1.172948	1.100630
3	colSums2(X[rows, cols])	2.007152	2.008952	1.955312	1.959815	1.985107	1.410043

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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	rowSums2_X_S	0.018653	0.0195865	0.0204666	0.0202965	0.0209870	0.034385
2	rowSums2(X, cols, rows)	0.020486	0.0221205	0.0231509	0.0228555	0.0235615	0.050862
3	rowSums2(X[cols, rows])	0.033788	0.0354585	0.0372226	0.0368025	0.0381555	0.052756

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.098268	1.129375	1.131156	1.126081	1.122671	1.479191
3	rowSums2(X[cols, rows])	1.811398	1.810354	1.818703	1.813244	1.818054	1.534274

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+10x1000 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_X_S() on double+10x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	17.197	17.7610	18.71455	18.4275	18.8005	39.869
2	rowSums2_X_S	18.653	19.5865	20.46658	20.2965	20.9870	34.385

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.0000	1.0000000
2	rowSums2_X_S	1.084666	1.102781	1.093619	1.101425	1.1163	0.8624495

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5290583 282.6    8529671 455.6  8529671 455.6
Vcells 10274172  78.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5290559 282.6    8529671 455.6  8529671 455.6
Vcells 10374225  79.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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	colSums2_X_S	0.082150	0.0891475	0.0984617	0.0958520	0.1064345	0.151089
2	colSums2(X, rows, cols)	0.098723	0.1078495	0.1185189	0.1152575	0.1288060	0.231682
3	colSums2(X[rows, cols])	0.184326	0.2030820	0.2248919	0.2203770	0.2389465	0.287071

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000
2	colSums2(X, rows, cols)	1.201741	1.209787	1.203705	1.202453	1.21019	1.533414
3	colSums2(X[rows, cols])	2.243774	2.278045	2.284054	2.299138	2.24501	1.900013

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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	rowSums2_X_S	0.095770	0.1044975	0.1169481	0.1134175	0.1237400	0.242178
2	rowSums2(X, cols, rows)	0.113031	0.1197200	0.1330936	0.1284435	0.1399495	0.273980
3	rowSums2(X[cols, rows])	0.197424	0.2228225	0.2452317	0.2382315	0.2621650	0.347064

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.180234	1.145673	1.138057	1.132484	1.130996	1.131317
3	rowSums2(X[cols, rows])	2.061439	2.132324	2.096928	2.100483	2.118676	1.433095

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+100x1000 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_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	colSums2_X_S	82.15	89.1475	98.46174	95.8520	106.4345	151.089
2	rowSums2_X_S	95.77	104.4975	116.94810	113.4175	123.7400	242.178

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2_X_S	1.165794	1.172187	1.187752	1.183256	1.162593	1.602883

Figure: Benchmarking of colSums2_X_S() and rowSums2_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  5290796 282.6    8529671 455.6  8529671 455.6
Vcells 10274316  78.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colSums2_X_S = colSums2(X_S, na.rm = FALSE), `colSums2(X, rows, cols)` = colSums2(X, 
+     rows = rows, cols = cols, na.rm = FALSE), `colSums2(X[rows, cols])` = colSums2(X[rows, cols], 
+     na.rm = FALSE), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5290772 282.6    8529671 455.6  8529671 455.6
Vcells 10374369  79.2   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowSums2_X_S = rowSums2(X_S, na.rm = FALSE), `rowSums2(X, cols, rows)` = rowSums2(X, 
+     rows = cols, cols = rows, na.rm = FALSE), `rowSums2(X[cols, rows])` = rowSums2(X[cols, rows], 
+     na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(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	colSums2_X_S	0.071943	0.0808235	0.0862944	0.0835675	0.0902290	0.13222
2	colSums2(X, rows, cols)	0.090710	0.1012370	0.1101146	0.1069040	0.1127460	0.15762
3	colSums2(X[rows, cols])	0.173366	0.1920180	0.2122097	0.2074070	0.2307825	0.35383

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colSums2(X, rows, cols)	1.260859	1.252569	1.276035	1.279253	1.249554	1.192104
3	colSums2(X[rows, cols])	2.409769	2.375769	2.459138	2.481910	2.557742	2.676070

Table: Benchmarking of rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(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	rowSums2_X_S	0.088471	0.0950455	0.1063684	0.104245	0.1147050	0.193939
2	rowSums2(X, cols, rows)	0.108314	0.1162975	0.1309940	0.126496	0.1359865	0.301330
3	rowSums2(X[cols, rows])	0.193556	0.2104240	0.2348916	0.231311	0.2525090	0.305308

	expr	min	lq	mean	median	uq	max
1	rowSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2(X, cols, rows)	1.224288	1.223598	1.231512	1.213449	1.185533	1.553736
3	rowSums2(X[cols, rows])	2.187790	2.213929	2.208283	2.218917	2.201377	1.574248

Figure: Benchmarking of colSums2_X_S(), colSums2(X, rows, cols)() and colSums2(X[rows, cols])() on double+1000x100 data as well as rowSums2_X_S(), rowSums2(X, cols, rows)() and rowSums2(X[cols, rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2_X_S() and rowSums2_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	colSums2_X_S	71.943	80.8235	86.29435	83.5675	90.229	132.220
2	rowSums2_X_S	88.471	95.0455	106.36845	104.2450	114.705	193.939

	expr	min	lq	mean	median	uq	max
1	colSums2_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
2	rowSums2_X_S	1.229737	1.175964	1.232624	1.247435	1.271265	1.46679

Figure: Benchmarking of colSums2_X_S() and rowSums2_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 23.11 secs.

Reproducibility

To reproduce this report, do:

html <- matrixStats:::benchmark('colRowSums2_subset')

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 19:09:06 (+0200 UTC). Powered by RSP.