matrixStats.benchmarks

---

colCumsums() and rowCumsums() benchmarks on subsetted computation

This report benchmark the performance of colCumsums() and rowCumsums() 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 5209099 278.2    8529671 455.6  8529671 455.6
Vcells 9847099  75.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5200736 277.8    8529671 455.6  8529671 455.6
Vcells 9819819  75.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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	colCumsums_X_S	0.001865	0.0020275	0.0029429	0.002125	0.0022135	0.082745
2	colCumsums(X, rows, cols)	0.002084	0.0023590	0.0025204	0.002461	0.0025720	0.004499
3	colCummins(X[rows, cols])	0.002808	0.0030800	0.0032182	0.003175	0.0032855	0.006142

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	colCumsums(X, rows, cols)	1.117426	1.163502	0.8564259	1.158118	1.161961	0.0543719
3	colCummins(X[rows, cols])	1.505630	1.519112	1.0935187	1.494118	1.484301	0.0742281

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums_X_S	0.001814	0.0021055	0.0022262	0.0021670	0.002280	0.004165
2	rowCumsums(X, cols, rows)	0.002228	0.0024165	0.0033511	0.0024905	0.002631	0.083943
3	rowCumsums(X[cols, rows])	0.002707	0.0031445	0.0032717	0.0032620	0.003332	0.005236

	expr	min	lq	mean	median	uq	max
1	rowCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums(X, cols, rows)	1.228225	1.147708	1.505307	1.149285	1.153947	20.154382
3	rowCumsums(X[cols, rows])	1.492282	1.493470	1.469637	1.505307	1.461404	1.257143

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	colCumsums_X_S	1.865	2.0275	2.94294	2.125	2.2135	82.745
2	rowCumsums_X_S	1.814	2.1055	2.22619	2.167	2.2800	4.165

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.0000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCumsums_X_S	0.9726542	1.038471	0.756451	1.019765	1.030043	0.0503354

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5199344 277.7    8529671 455.6  8529671 455.6
Vcells 9488596  72.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5199320 277.7    8529671 455.6  8529671 455.6
Vcells 9493649  72.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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
2	colCumsums(X, rows, cols)	0.019051	0.0201260	0.0206943	0.0207345	0.0210590	0.025470
1	colCumsums_X_S	0.020088	0.0210365	0.0217494	0.0218750	0.0220415	0.035294
3	colCummins(X[rows, cols])	0.030387	0.0317610	0.0330226	0.0329485	0.0334765	0.062142

	expr	min	lq	mean	median	uq	max
2	colCumsums(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCumsums_X_S	1.054433	1.045240	1.050981	1.055005	1.046655	1.385709
3	colCummins(X[rows, cols])	1.595034	1.578108	1.595733	1.589066	1.589653	2.439811

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums(X, cols, rows)	0.020352	0.021432	0.0223430	0.0220085	0.0223570	0.049823
1	rowCumsums_X_S	0.021351	0.022402	0.0234976	0.0232995	0.0240555	0.037651
3	rowCumsums(X[cols, rows])	0.031482	0.032849	0.0340993	0.0341365	0.0344790	0.049689

	expr	min	lq	mean	median	uq	max
2	rowCumsums(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowCumsums_X_S	1.049086	1.045259	1.051674	1.058659	1.075972	0.7556952
3	rowCumsums(X[cols, rows])	1.546875	1.532708	1.526172	1.551060	1.542202	0.9973105

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	colCumsums_X_S	20.088	21.0365	21.74935	21.8750	22.0415	35.294
2	rowCumsums_X_S	21.351	22.4020	23.49760	23.2995	24.0555	37.651

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000
2	rowCumsums_X_S	1.062873	1.064911	1.080382	1.06512	1.091373	1.066782

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5200086 277.8    8529671 455.6  8529671 455.6
Vcells 9492647  72.5   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5200062 277.8    8529671 455.6  8529671 455.6
Vcells 9497700  72.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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	colCumsums_X_S	0.019896	0.0209990	0.0228349	0.0211610	0.0216220	0.054841
2	colCumsums(X, rows, cols)	0.021220	0.0222705	0.0250813	0.0226300	0.0233790	0.042252
3	colCummins(X[rows, cols])	0.031564	0.0328660	0.0359631	0.0332225	0.0339785	0.068507

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colCumsums(X, rows, cols)	1.066546	1.060550	1.098374	1.069420	1.081260	0.7704455
3	colCummins(X[rows, cols])	1.586450	1.565122	1.574917	1.569987	1.571478	1.2491931

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums_X_S	0.023157	0.0241025	0.0250593	0.0249085	0.0253485	0.039821
2	rowCumsums(X, cols, rows)	0.024115	0.0253970	0.0266695	0.0265325	0.0271485	0.041174
3	rowCumsums(X[cols, rows])	0.035286	0.0368020	0.0384760	0.0382340	0.0388655	0.070778

	expr	min	lq	mean	median	uq	max
1	rowCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums(X, cols, rows)	1.041370	1.053708	1.064254	1.065199	1.071010	1.033977
3	rowCumsums(X[cols, rows])	1.523772	1.526896	1.535397	1.534978	1.533246	1.777404

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	colCumsums_X_S	19.896	20.9990	22.83492	21.1610	21.6220	54.841
2	rowCumsums_X_S	23.157	24.1025	25.05931	24.9085	25.3485	39.821

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCumsums_X_S	1.163902	1.147793	1.097412	1.177095	1.172348	0.7261173

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5200291 277.8    8529671 455.6  8529671 455.6
Vcells 9493478  72.5   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5200267 277.8    8529671 455.6  8529671 455.6
Vcells 9498531  72.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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	colCumsums_X_S	0.020663	0.0217285	0.0228216	0.0225240	0.0227875	0.039375
2	colCumsums(X, rows, cols)	0.022618	0.0234350	0.0243940	0.0238195	0.0246235	0.041735
3	colCummins(X[rows, cols])	0.032926	0.0343930	0.0356181	0.0355960	0.0360740	0.049914

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCumsums(X, rows, cols)	1.094614	1.078537	1.068899	1.057516	1.080571	1.059937
3	colCummins(X[rows, cols])	1.593476	1.582852	1.560716	1.580359	1.583061	1.267657

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums_X_S	0.020002	0.0208975	0.0214689	0.0210570	0.0218395	0.032830
2	rowCumsums(X, cols, rows)	0.020517	0.0213845	0.0225733	0.0221090	0.0227570	0.050701
3	rowCumsums(X[cols, rows])	0.030428	0.0316715	0.0327162	0.0321275	0.0334550	0.051402

	expr	min	lq	mean	median	uq	max
1	rowCumsums_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000
2	rowCumsums(X, cols, rows)	1.025747	1.023304	1.051438	1.04996	1.042011	1.544350
3	rowCumsums(X[cols, rows])	1.521248	1.515564	1.523885	1.52574	1.531857	1.565702

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	rowCumsums_X_S	20.002	20.8975	21.46894	21.057	21.8395	32.830
1	colCumsums_X_S	20.663	21.7285	22.82161	22.524	22.7875	39.375

	expr	min	lq	mean	median	uq	max
2	rowCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
1	colCumsums_X_S	1.033047	1.039765	1.063006	1.069668	1.043408	1.19936

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5200502 277.8    8529671 455.6  8529671 455.6
Vcells 9516151  72.7   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5200478 277.8    8529671 455.6  8529671 455.6
Vcells 9566204  73.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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
2	colCumsums(X, rows, cols)	0.114622	0.1243840	0.1399265	0.1357755	0.1505505	0.250230
1	colCumsums_X_S	0.120870	0.1311700	0.1469682	0.1432735	0.1580940	0.206596
3	colCummins(X[rows, cols])	0.187301	0.2031185	0.2304772	0.2286825	0.2467070	0.307835

	expr	min	lq	mean	median	uq	max
2	colCumsums(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colCumsums_X_S	1.054510	1.054557	1.050324	1.055224	1.050106	0.8256244
3	colCummins(X[rows, cols])	1.634076	1.632995	1.647131	1.684269	1.638699	1.2302082

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums(X, cols, rows)	0.121574	0.1286955	0.1447632	0.1363945	0.156258	0.255207
1	rowCumsums_X_S	0.128468	0.1398965	0.1607373	0.1581045	0.173563	0.230483
3	rowCumsums(X[cols, rows])	0.191760	0.2062265	0.2350259	0.2276565	0.257936	0.329617

	expr	min	lq	mean	median	uq	max
2	rowCumsums(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowCumsums_X_S	1.056706	1.087035	1.110347	1.159171	1.110746	0.9031218
3	rowCumsums(X[cols, rows])	1.577311	1.602437	1.623520	1.669103	1.650706	1.2915672

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	colCumsums_X_S	120.870	131.1700	146.9682	143.2735	158.094	206.596
2	rowCumsums_X_S	128.468	139.8965	160.7373	158.1045	173.563	230.483

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums_X_S	1.062861	1.066528	1.093688	1.103515	1.097847	1.115622

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5200712 277.8    8529671 455.6  8529671 455.6
Vcells 9516926  72.7   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5200688 277.8    8529671 455.6  8529671 455.6
Vcells 9566979  73.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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
2	colCumsums(X, rows, cols)	0.111316	0.1173475	0.1331221	0.1320605	0.1413450	0.180476
1	colCumsums_X_S	0.116174	0.1224300	0.1359007	0.1333120	0.1468025	0.187023
3	colCummins(X[rows, cols])	0.179700	0.1878335	0.2151261	0.2136050	0.2347925	0.316732

	expr	min	lq	mean	median	uq	max
2	colCumsums(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCumsums_X_S	1.043642	1.043311	1.020873	1.009477	1.038611	1.036276
3	colCummins(X[rows, cols])	1.614323	1.600660	1.616005	1.617478	1.661131	1.754981

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums(X, cols, rows)	0.125494	0.1360550	0.1535297	0.1488635	0.1639980	0.268564
1	rowCumsums_X_S	0.134486	0.1415375	0.1619960	0.1592740	0.1754235	0.222288
3	rowCumsums(X[cols, rows])	0.201314	0.2130210	0.2425892	0.2380050	0.2651395	0.324543

	expr	min	lq	mean	median	uq	max
2	rowCumsums(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	rowCumsums_X_S	1.071653	1.040296	1.055144	1.069933	1.069668	0.827691
3	rowCumsums(X[cols, rows])	1.604172	1.565698	1.580080	1.598814	1.616724	1.208438

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	colCumsums_X_S	116.174	122.4300	135.9007	133.312	146.8025	187.023
2	rowCumsums_X_S	134.486	141.5375	161.9960	159.274	175.4235	222.288

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
2	rowCumsums_X_S	1.157626	1.156069	1.192017	1.194746	1.194963	1.18856

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5200930 277.8    8529671 455.6  8529671 455.6
Vcells 9608033  73.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5200897 277.8    8529671 455.6  8529671 455.6
Vcells 9608171  73.4   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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	colCumsums_X_S	0.001788	0.001956	0.0022014	0.0020330	0.0021255	0.016534
2	colCumsums(X, rows, cols)	0.002103	0.002274	0.0023992	0.0023325	0.0024335	0.004869
3	colCummins(X[rows, cols])	0.002886	0.003062	0.0032185	0.0031600	0.0032640	0.008024

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colCumsums(X, rows, cols)	1.176174	1.162577	1.089866	1.147319	1.144907	0.2944841
3	colCummins(X[rows, cols])	1.614094	1.565440	1.462026	1.554353	1.535639	0.4853030

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums_X_S	0.001740	0.0018990	0.0020286	0.001971	0.0020695	0.003978
2	rowCumsums(X, cols, rows)	0.002095	0.0022325	0.0025121	0.002294	0.0024070	0.019733
3	rowCumsums(X[cols, rows])	0.002707	0.0029685	0.0031006	0.003056	0.0031865	0.005149

	expr	min	lq	mean	median	uq	max
1	rowCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums(X, cols, rows)	1.204023	1.175619	1.238361	1.163876	1.163083	4.960533
3	rowCumsums(X[cols, rows])	1.555747	1.563191	1.528419	1.550482	1.539744	1.294369

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

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

	expr	min	lq	mean	median	uq	max
2	rowCumsums_X_S	1.740	1.899	2.02860	1.971	2.0695	3.978
1	colCumsums_X_S	1.788	1.956	2.20139	2.033	2.1255	16.534

	expr	min	lq	mean	median	uq	max
2	rowCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.00000	1.00000
1	colCumsums_X_S	1.027586	1.030016	1.085177	1.031456	1.02706	4.15636

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5201129 277.8    8529671 455.6  8529671 455.6
Vcells 9613984  73.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5201105 277.8    8529671 455.6  8529671 455.6
Vcells 9624037  73.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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	colCumsums(X, rows, cols)	0.012510	0.0131305	0.0137650	0.0136055	0.0139070	0.022369
1	colCumsums_X_S	0.014541	0.0151970	0.0155555	0.0154435	0.0156120	0.029622
3	colCummins(X[rows, cols])	0.029763	0.0305910	0.0314978	0.0312000	0.0313975	0.057288

	expr	min	lq	mean	median	uq	max
2	colCumsums(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCumsums_X_S	1.162350	1.157382	1.130081	1.135092	1.122600	1.324243
3	colCummins(X[rows, cols])	2.379137	2.329767	2.288257	2.293190	2.257676	2.561044

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums_X_S	0.015987	0.0165495	0.0170890	0.0171005	0.0173515	0.021495
2	rowCumsums(X, cols, rows)	0.016578	0.0171580	0.0184190	0.0176000	0.0180890	0.042720
3	rowCumsums(X[cols, rows])	0.031220	0.0318270	0.0328658	0.0328910	0.0331765	0.047805

	expr	min	lq	mean	median	uq	max
1	rowCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums(X, cols, rows)	1.036968	1.036768	1.077824	1.029210	1.042504	1.987439
3	rowCumsums(X[cols, rows])	1.952837	1.923140	1.923210	1.923394	1.912025	2.224006

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	colCumsums_X_S	14.541	15.1970	15.55554	15.4435	15.6120	29.622
2	rowCumsums_X_S	15.987	16.5495	17.08904	17.1005	17.3515	21.495

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowCumsums_X_S	1.099443	1.088998	1.098582	1.107294	1.111421	0.7256431

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5201328 277.8    8529671 455.6  8529671 455.6
Vcells 9615384  73.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5201304 277.8    8529671 455.6  8529671 455.6
Vcells 9625437  73.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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	colCumsums_X_S	0.013636	0.0142150	0.0146778	0.0143670	0.014572	0.028476
2	colCumsums(X, rows, cols)	0.013671	0.0145220	0.0149880	0.0148885	0.015318	0.019313
3	colCummins(X[rows, cols])	0.029080	0.0298125	0.0310441	0.0306755	0.030983	0.057977

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colCumsums(X, rows, cols)	1.002567	1.021597	1.021137	1.036299	1.051194	0.6782203
3	colCummins(X[rows, cols])	2.132590	2.097256	2.115042	2.135136	2.126201	2.0359952

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums_X_S	0.017414	0.0179620	0.0186248	0.0181975	0.0188450	0.030882
2	rowCumsums(X, cols, rows)	0.020751	0.0215475	0.0224733	0.0221870	0.0225515	0.035438
3	rowCumsums(X[cols, rows])	0.035460	0.0363805	0.0379843	0.0367905	0.0380045	0.070058

	expr	min	lq	mean	median	uq	max
1	rowCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums(X, cols, rows)	1.191627	1.199616	1.206636	1.219233	1.196684	1.147529
3	rowCumsums(X[cols, rows])	2.036293	2.025415	2.039450	2.021734	2.016689	2.268571

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	colCumsums_X_S	13.636	14.215	14.67779	14.3670	14.572	28.476
2	rowCumsums_X_S	17.414	17.962	18.62477	18.1975	18.845	30.882

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums_X_S	1.277061	1.263595	1.268908	1.266618	1.293234	1.084492

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5201533 277.8    8529671 455.6  8529671 455.6
Vcells 9615520  73.4   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5201509 277.8    8529671 455.6  8529671 455.6
Vcells 9625573  73.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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	colCumsums_X_S	0.015375	0.0160580	0.0165749	0.0164700	0.0168435	0.031113
2	colCumsums(X, rows, cols)	0.015413	0.0163180	0.0173903	0.0167470	0.0175030	0.036887
3	colCummins(X[rows, cols])	0.033465	0.0343655	0.0354200	0.0350205	0.0359240	0.049779

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colCumsums(X, rows, cols)	1.002471	1.016191	1.049194	1.016819	1.039155	1.185582
3	colCummins(X[rows, cols])	2.176585	2.140086	2.136965	2.126321	2.132811	1.599942

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums_X_S	0.014709	0.015229	0.0157979	0.0155925	0.0158990	0.029984
2	rowCumsums(X, cols, rows)	0.017294	0.017991	0.0189216	0.0185060	0.0189190	0.044872
3	rowCumsums(X[cols, rows])	0.030251	0.031014	0.0322598	0.0320335	0.0323365	0.048089

	expr	min	lq	mean	median	uq	max
1	rowCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums(X, cols, rows)	1.175743	1.181364	1.197728	1.186853	1.189949	1.496531
3	rowCumsums(X[cols, rows])	2.056632	2.036509	2.042023	2.054417	2.033870	1.603822

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	rowCumsums_X_S	14.709	15.229	15.79794	15.5925	15.8990	29.984
1	colCumsums_X_S	15.375	16.058	16.57490	16.4700	16.8435	31.113

	expr	min	lq	mean	median	uq	max
2	rowCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colCumsums_X_S	1.045278	1.054436	1.049181	1.056277	1.059406	1.037653

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5201744 277.9    8529671 455.6  8529671 455.6
Vcells 9660975  73.8   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5201720 277.9    8529671 455.6  8529671 455.6
Vcells 9761028  74.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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
2	colCumsums(X, rows, cols)	0.078123	0.0852240	0.0948696	0.0910215	0.0987835	0.222979
1	colCumsums_X_S	0.089342	0.0974465	0.1065202	0.1038410	0.1125170	0.157940
3	colCummins(X[rows, cols])	0.190537	0.2106785	0.2312126	0.2282060	0.2427230	0.421410

	expr	min	lq	mean	median	uq	max
2	colCumsums(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colCumsums_X_S	1.143607	1.143416	1.122807	1.140840	1.139026	0.7083178
3	colCummins(X[rows, cols])	2.438936	2.472056	2.437163	2.507166	2.457121	1.8899089

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums(X, cols, rows)	0.098962	0.1045880	0.1169032	0.1111255	0.1249150	0.247900
1	rowCumsums_X_S	0.094117	0.1057070	0.1172707	0.1149435	0.1248715	0.167960
3	rowCumsums(X[cols, rows])	0.195717	0.2139035	0.2352298	0.2326945	0.2493180	0.345397

	expr	min	lq	mean	median	uq	max
2	rowCumsums(X, cols, rows)	1.0000000	1.000000	1.000000	1.000000	1.0000000	1.0000000
1	rowCumsums_X_S	0.9510418	1.010699	1.003143	1.034358	0.9996518	0.6775313
3	rowCumsums(X[cols, rows])	1.9776985	2.045201	2.012176	2.093979	1.9959012	1.3932916

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	colCumsums_X_S	89.342	97.4465	106.5202	103.8410	112.5170	157.94
2	rowCumsums_X_S	94.117	105.7070	117.2707	114.9435	124.8715	167.96

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums_X_S	1.053446	1.08477	1.100925	1.106918	1.109801	1.063442

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 5201954 277.9    8529671 455.6  8529671 455.6
Vcells 9661116  73.8   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colCumsums_X_S = colCumsums(X_S), `colCumsums(X, rows, cols)` = colCumsums(X, 
+     rows = rows, cols = cols), `colCummins(X[rows, cols])` = colCumsums(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5201930 277.9    8529671 455.6  8529671 455.6
Vcells 9761169  74.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowCumsums_X_S = rowCumsums(X_S), `rowCumsums(X, cols, rows)` = rowCumsums(X, 
+     rows = cols, cols = rows), `rowCumsums(X[cols, rows])` = rowCumsums(X[cols, rows]), unit = "ms")

Table: Benchmarking of colCumsums_X_S(), colCumsums(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	colCumsums(X, rows, cols)	0.077105	0.0818540	0.0912960	0.0895540	0.0941125	0.188140
1	colCumsums_X_S	0.087339	0.0926745	0.1017743	0.0989260	0.1063610	0.147469
3	colCummins(X[rows, cols])	0.192694	0.2058950	0.2269163	0.2252045	0.2423040	0.384437

	expr	min	lq	mean	median	uq	max
2	colCumsums(X, rows, cols)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colCumsums_X_S	1.132728	1.132193	1.114774	1.104652	1.130147	0.7838259
3	colCummins(X[rows, cols])	2.499112	2.515393	2.485502	2.514734	2.574621	2.0433560

Table: Benchmarking of rowCumsums_X_S(), rowCumsums(X, cols, rows)() and rowCumsums(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	rowCumsums_X_S	0.099372	0.1077550	0.1208908	0.1183805	0.1279110	0.168878
2	rowCumsums(X, cols, rows)	0.097971	0.1070255	0.1214668	0.1201700	0.1275320	0.262877
3	rowCumsums(X[cols, rows])	0.197242	0.2158110	0.2400602	0.2352330	0.2568715	0.322989

	expr	min	lq	mean	median	uq	max
1	rowCumsums_X_S	1.0000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums(X, cols, rows)	0.9859015	0.993230	1.004765	1.015116	0.997037	1.556609
3	rowCumsums(X[cols, rows])	1.9848851	2.002793	1.985761	1.987092	2.008205	1.912558

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

Table: Benchmarking of colCumsums_X_S() and rowCumsums_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	colCumsums_X_S	87.339	92.6745	101.7743	98.9260	106.361	147.469
2	rowCumsums_X_S	99.372	107.7550	120.8908	118.3805	127.911	168.878

	expr	min	lq	mean	median	uq	max
1	colCumsums_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowCumsums_X_S	1.137773	1.162726	1.187831	1.196657	1.202612	1.145176

Figure: Benchmarking of colCumsums_X_S() and rowCumsums_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 21.85 secs.

Reproducibility

To reproduce this report, do:

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

---

Copyright Dongcan Jiang. Last updated on 2021-08-25 18:54:20 (+0200 UTC). Powered by RSP.