matrixStats.benchmarks

---

colDiffs() and rowDiffs() benchmarks on subsetted computation

This report benchmark the performance of colDiffs() and rowDiffs() 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 5216753 278.7    8529671 455.6  8529671 455.6
Vcells 9900928  75.6   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5208390 278.2    8529671 455.6  8529671 455.6
Vcells 9873644  75.4   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

Table: Benchmarking of colDiffs_X_S(), colDiffs(X, rows, cols)() and colDiffs(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	colDiffs_X_S	0.001853	0.0020885	0.0030051	0.0021650	0.0022745	0.081957
2	colDiffs(X, rows, cols)	0.002282	0.0025300	0.0026796	0.0026200	0.0027160	0.005247
3	colDiffs(X[rows, cols])	0.002682	0.0030960	0.0033030	0.0032005	0.0033445	0.006976

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	colDiffs(X, rows, cols)	1.231516	1.211396	0.8916689	1.210162	1.194109	0.0640214
3	colDiffs(X[rows, cols])	1.447383	1.482404	1.0991069	1.478291	1.470433	0.0851178

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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	rowDiffs_X_S	0.002081	0.0022595	0.0024620	0.0023425	0.0024795	0.004759
2	rowDiffs(X, cols, rows)	0.002440	0.0026090	0.0037223	0.0027035	0.0028260	0.087765
3	rowDiffs(X[cols, rows])	0.002922	0.0032600	0.0034573	0.0033540	0.0035450	0.006753

	expr	min	lq	mean	median	uq	max
1	rowDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs(X, cols, rows)	1.172513	1.154680	1.511928	1.154109	1.139746	18.441900
3	rowDiffs(X[cols, rows])	1.404133	1.442797	1.404293	1.431804	1.429724	1.418996

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	1.853	2.0885	3.00514	2.1650	2.2745	81.957
2	rowDiffs_X_S	2.081	2.2595	2.46195	2.3425	2.4795	4.759

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.0000000	1.000000	1.00000	1.000000
2	rowDiffs_X_S	1.123044	1.081877	0.8192464	1.081986	1.09013	0.058067

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5207004 278.1    8529671 455.6  8529671 455.6
Vcells 9542513  72.9   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5206980 278.1    8529671 455.6  8529671 455.6
Vcells 9547566  72.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	0.016574	0.0173745	0.0180267	0.0179960	0.0182000	0.031714
2	colDiffs(X, rows, cols)	0.020271	0.0212740	0.0218350	0.0218275	0.0222450	0.026163
3	colDiffs(X[rows, cols])	0.026540	0.0277900	0.0289398	0.0288405	0.0292585	0.056312

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colDiffs(X, rows, cols)	1.223060	1.224438	1.211256	1.212908	1.222253	0.8249669
3	colDiffs(X[rows, cols])	1.601303	1.599471	1.605381	1.602606	1.607610	1.7756196

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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	rowDiffs(X, cols, rows)	0.023333	0.0244220	0.0256964	0.0253795	0.025925	0.055503
1	rowDiffs_X_S	0.024787	0.0259895	0.0272196	0.0271065	0.028147	0.044725
3	rowDiffs(X[cols, rows])	0.034909	0.0364875	0.0380877	0.0379805	0.038617	0.054645

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowDiffs_X_S	1.062315	1.064184	1.059278	1.068047	1.085709	0.8058123
3	rowDiffs(X[cols, rows])	1.496121	1.494042	1.482218	1.496503	1.489566	0.9845414

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	16.574	17.3745	18.02672	17.9960	18.200	31.714
2	rowDiffs_X_S	24.787	25.9895	27.21962	27.1065	28.147	44.725

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.00000	1.000000	1.000000	1.00000
2	rowDiffs_X_S	1.495535	1.495842	1.50996	1.506251	1.546539	1.41026

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5207746 278.2    8529671 455.6  8529671 455.6
Vcells 9546579  72.9   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5207722 278.2    8529671 455.6  8529671 455.6
Vcells 9551632  72.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

Table: Benchmarking of colDiffs_X_S(), colDiffs(X, rows, cols)() and colDiffs(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	colDiffs_X_S	0.016924	0.017648	0.0182821	0.0178145	0.0184195	0.032679
2	colDiffs(X, rows, cols)	0.021740	0.022522	0.0231821	0.0228625	0.0234785	0.034067
3	colDiffs(X[rows, cols])	0.027479	0.028699	0.0297399	0.0289215	0.0299665	0.058360

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colDiffs(X, rows, cols)	1.284566	1.276179	1.268025	1.283365	1.274655	1.042474
3	colDiffs(X[rows, cols])	1.623671	1.626190	1.626728	1.623481	1.626890	1.785856

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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	rowDiffs_X_S	0.025358	0.0263175	0.0272596	0.0271155	0.0283135	0.032790
2	rowDiffs(X, cols, rows)	0.027049	0.0275110	0.0286743	0.0284785	0.0296440	0.035178
3	rowDiffs(X[cols, rows])	0.037066	0.0388010	0.0409239	0.0401090	0.0418575	0.072411

	expr	min	lq	mean	median	uq	max
1	rowDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs(X, cols, rows)	1.066685	1.045350	1.051896	1.050266	1.046992	1.072827
3	rowDiffs(X[cols, rows])	1.461708	1.474342	1.501262	1.479191	1.478358	2.208326

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	16.924	17.6480	18.28206	17.8145	18.4195	32.679
2	rowDiffs_X_S	25.358	26.3175	27.25965	27.1155	28.3135	32.790

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.00000	1.000000	1.000000	1.000000
2	rowDiffs_X_S	1.498346	1.491246	1.49106	1.522103	1.537148	1.003397

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5207951 278.2    8529671 455.6  8529671 455.6
Vcells 9547383  72.9   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5207927 278.2    8529671 455.6  8529671 455.6
Vcells 9552436  72.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

Table: Benchmarking of colDiffs_X_S(), colDiffs(X, rows, cols)() and colDiffs(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	colDiffs_X_S	0.015870	0.0167690	0.0176202	0.0172355	0.0175295	0.033904
2	colDiffs(X, rows, cols)	0.019952	0.0211185	0.0220102	0.0213615	0.0221685	0.037966
3	colDiffs(X[rows, cols])	0.028584	0.0299000	0.0311677	0.0308900	0.0313985	0.044057

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colDiffs(X, rows, cols)	1.257215	1.259377	1.249148	1.239390	1.264640	1.119809
3	colDiffs(X[rows, cols])	1.801134	1.783052	1.768866	1.792231	1.791181	1.299463

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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	rowDiffs_X_S	0.021083	0.021359	0.0224119	0.0221535	0.0230205	0.028383
2	rowDiffs(X, cols, rows)	0.022123	0.022477	0.0234689	0.0230855	0.0238865	0.046887
3	rowDiffs(X[cols, rows])	0.031814	0.032152	0.0339521	0.0333495	0.0347535	0.049268

	expr	min	lq	mean	median	uq	max
1	rowDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs(X, cols, rows)	1.049329	1.052343	1.047164	1.042070	1.037619	1.651939
3	rowDiffs(X[cols, rows])	1.508988	1.505314	1.514914	1.505383	1.509676	1.735828

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

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

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	15.870	16.769	17.62017	17.2355	17.5295	33.904
2	rowDiffs_X_S	21.083	21.359	22.41188	22.1535	23.0205	28.383

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowDiffs_X_S	1.328481	1.273719	1.271945	1.285341	1.313243	0.8371579

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5208162 278.2    8529671 455.6  8529671 455.6
Vcells 9570047  73.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5208138 278.2    8529671 455.6  8529671 455.6
Vcells 9620100  73.4   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	0.105206	0.1141290	0.1273333	0.1250335	0.1339815	0.185541
2	colDiffs(X, rows, cols)	0.126209	0.1374505	0.1539970	0.1503190	0.1659385	0.271798
3	colDiffs(X[rows, cols])	0.173455	0.1886770	0.2116374	0.2124520	0.2262440	0.274103

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colDiffs(X, rows, cols)	1.199637	1.204343	1.209401	1.202230	1.238518	1.464895
3	colDiffs(X[rows, cols])	1.648718	1.653191	1.662074	1.699161	1.688621	1.477318

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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	rowDiffs(X, cols, rows)	0.141634	0.143431	0.1641314	0.1540035	0.1791340	0.293898
1	rowDiffs_X_S	0.151821	0.160112	0.1849618	0.1806230	0.1996705	0.273409
3	rowDiffs(X[cols, rows])	0.215160	0.220846	0.2567680	0.2489070	0.2800360	0.359153

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowDiffs_X_S	1.071925	1.116300	1.126913	1.172850	1.114643	0.9302853
3	rowDiffs(X[cols, rows])	1.519127	1.539737	1.564406	1.616243	1.563277	1.2220328

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	105.206	114.129	127.3333	125.0335	133.9815	185.541
2	rowDiffs_X_S	151.821	160.112	184.9617	180.6230	199.6705	273.409

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs_X_S	1.443083	1.402904	1.452579	1.444597	1.490284	1.473577

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5208372 278.2    8529671 455.6  8529671 455.6
Vcells 9570765  73.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5208348 278.2    8529671 455.6  8529671 455.6
Vcells 9620818  73.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

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

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	0.103472	0.1137520	0.1272180	0.1230740	0.1346685	0.218679
2	colDiffs(X, rows, cols)	0.122819	0.1335395	0.1501142	0.1504485	0.1582405	0.204807
3	colDiffs(X[rows, cols])	0.167456	0.1864720	0.2100959	0.2098080	0.2318155	0.329141

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colDiffs(X, rows, cols)	1.186978	1.173953	1.179977	1.222423	1.175037	0.9365646
3	colDiffs(X[rows, cols])	1.618370	1.639285	1.651464	1.704730	1.721379	1.5051331

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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	rowDiffs(X, cols, rows)	0.144292	0.1488515	0.1721450	0.1660545	0.1877765	0.299767
1	rowDiffs_X_S	0.158597	0.1617020	0.1892060	0.1826615	0.2071985	0.274642
3	rowDiffs(X[cols, rows])	0.227369	0.2306220	0.2682557	0.2549265	0.2965480	0.368254

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowDiffs_X_S	1.099139	1.086331	1.099108	1.100009	1.103431	0.9161849
3	rowDiffs(X[cols, rows])	1.575756	1.549343	1.558313	1.535198	1.579260	1.2284674

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	103.472	113.752	127.218	123.0740	134.6685	218.679
2	rowDiffs_X_S	158.597	161.702	189.206	182.6615	207.1985	274.642

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000
2	rowDiffs_X_S	1.532753	1.421531	1.487258	1.48416	1.538582	1.255914

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5208590 278.2    8529671 455.6  8529671 455.6
Vcells 9661854  73.8   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5208557 278.2    8529671 455.6  8529671 455.6
Vcells 9661992  73.8   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

Table: Benchmarking of colDiffs_X_S(), colDiffs(X, rows, cols)() and colDiffs(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	colDiffs_X_S	0.002027	0.002165	0.0023908	0.002231	0.002321	0.015429
2	colDiffs(X, rows, cols)	0.002301	0.002575	0.0027069	0.002640	0.002741	0.005425
3	colDiffs(X[rows, cols])	0.002939	0.003178	0.0033462	0.003304	0.003390	0.007236

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colDiffs(X, rows, cols)	1.135175	1.189376	1.132234	1.183326	1.180956	0.3516106
3	colDiffs(X[rows, cols])	1.449926	1.467898	1.399639	1.480950	1.460577	0.4689870

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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	rowDiffs_X_S	0.002000	0.0022050	0.0023352	0.002259	0.0023600	0.004508
2	rowDiffs(X, cols, rows)	0.002339	0.0025395	0.0028191	0.002617	0.0027345	0.019367
3	rowDiffs(X[cols, rows])	0.002918	0.0032200	0.0033761	0.003306	0.0034385	0.006170

	expr	min	lq	mean	median	uq	max
1	rowDiffs_X_S	1.0000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs(X, cols, rows)	1.1695	1.151701	1.207223	1.158477	1.158686	4.296140
3	rowDiffs(X[cols, rows])	1.4590	1.460317	1.445750	1.463479	1.456991	1.368678

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	2.027	2.165	2.39076	2.231	2.321	15.429
2	rowDiffs_X_S	2.000	2.205	2.33521	2.259	2.360	4.508

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.0000000	1.000000	1.0000000	1.00000	1.000000	1.0000000
2	rowDiffs_X_S	0.9866798	1.018476	0.9767647	1.01255	1.016803	0.2921771

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5208789 278.2    8529671 455.6  8529671 455.6
Vcells 9667788  73.8   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5208765 278.2    8529671 455.6  8529671 455.6
Vcells 9677841  73.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

Table: Benchmarking of colDiffs_X_S(), colDiffs(X, rows, cols)() and colDiffs(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	colDiffs_X_S	0.017660	0.0182725	0.0192073	0.018885	0.0194270	0.032789
2	colDiffs(X, rows, cols)	0.021828	0.0226875	0.0233230	0.023063	0.0236800	0.028592
3	colDiffs(X[rows, cols])	0.032381	0.0334315	0.0346890	0.033991	0.0350635	0.060921

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colDiffs(X, rows, cols)	1.236014	1.241620	1.214279	1.221234	1.218922	0.8719998
3	colDiffs(X[rows, cols])	1.833579	1.829607	1.806029	1.799894	1.804885	1.8579707

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

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	0.018192	0.0191005	0.0200181	0.0194705	0.0198285	0.046732
1	rowDiffs_X_S	0.021486	0.0221085	0.0228331	0.0223885	0.0231845	0.037368
3	rowDiffs(X[cols, rows])	0.035955	0.0373810	0.0384471	0.0377105	0.0389640	0.051352

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowDiffs_X_S	1.181069	1.157483	1.140620	1.149868	1.169251	0.7996234
3	rowDiffs(X[cols, rows])	1.976418	1.957069	1.920613	1.936802	1.965050	1.0988616

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	17.660	18.2725	19.20731	18.8850	19.4270	32.789
2	rowDiffs_X_S	21.486	22.1085	22.83310	22.3885	23.1845	37.368

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
2	rowDiffs_X_S	1.216648	1.209933	1.188771	1.185518	1.193416	1.13965

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5208988 278.2    8529671 455.6  8529671 455.6
Vcells 9669163  73.8   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5208964 278.2    8529671 455.6  8529671 455.6
Vcells 9679216  73.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

Table: Benchmarking of colDiffs_X_S(), colDiffs(X, rows, cols)() and colDiffs(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	colDiffs_X_S	0.017071	0.0177390	0.0182864	0.0181515	0.0186175	0.023403
2	colDiffs(X, rows, cols)	0.023282	0.0242775	0.0252079	0.0246955	0.0255510	0.038888
3	colDiffs(X[rows, cols])	0.032231	0.0333675	0.0353652	0.0342595	0.0349865	0.065132

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colDiffs(X, rows, cols)	1.363833	1.368595	1.378502	1.360521	1.372418	1.661667
3	colDiffs(X[rows, cols])	1.888056	1.881025	1.933956	1.887420	1.879226	2.783062

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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
2	rowDiffs(X, cols, rows)	0.021593	0.0224475	0.0233552	0.0232645	0.0237035	0.035781
1	rowDiffs_X_S	0.022180	0.0231025	0.0241712	0.0238160	0.0247260	0.038145
3	rowDiffs(X[cols, rows])	0.039317	0.0408925	0.0429012	0.0422995	0.0438610	0.076166

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	rowDiffs_X_S	1.027185	1.029179	1.034937	1.023706	1.043137	1.066069
3	rowDiffs(X[cols, rows])	1.820822	1.821695	1.836902	1.818199	1.850402	2.128672

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	17.071	17.7390	18.28645	18.1515	18.6175	23.403
2	rowDiffs_X_S	22.180	23.1025	24.17117	23.8160	24.7260	38.145

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs_X_S	1.299279	1.302356	1.321808	1.312068	1.328105	1.629919

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5209193 278.3    8529671 455.6  8529671 455.6
Vcells 9669299  73.8   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5209169 278.2    8529671 455.6  8529671 455.6
Vcells 9679352  73.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

Table: Benchmarking of colDiffs_X_S(), colDiffs(X, rows, cols)() and colDiffs(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	colDiffs_X_S	0.017365	0.0181230	0.0190861	0.0190085	0.0192945	0.036127
2	colDiffs(X, rows, cols)	0.021960	0.0225610	0.0237898	0.0235120	0.0241440	0.041742
3	colDiffs(X[rows, cols])	0.035284	0.0369825	0.0384188	0.0381935	0.0384775	0.061739

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colDiffs(X, rows, cols)	1.264613	1.244882	1.246449	1.236920	1.251341	1.155424
3	colDiffs(X[rows, cols])	2.031903	2.040639	2.012926	2.009285	1.994221	1.708943

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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
2	rowDiffs(X, cols, rows)	0.017762	0.018901	0.0196780	0.019386	0.0197765	0.045941
1	rowDiffs_X_S	0.016350	0.019423	0.0198705	0.019788	0.0200795	0.034849
3	rowDiffs(X[cols, rows])	0.033709	0.035291	0.0361828	0.036015	0.0364650	0.049748

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	1.0000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowDiffs_X_S	0.9205044	1.027618	1.009785	1.020737	1.015321	0.7585599
3	rowDiffs(X[cols, rows])	1.8978156	1.867150	1.838743	1.857784	1.843855	1.0828672

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	17.365	18.123	19.08606	19.0085	19.2945	36.127
2	rowDiffs_X_S	16.350	19.423	19.87054	19.7880	20.0795	34.849

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.0000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowDiffs_X_S	0.9415491	1.071732	1.041102	1.041008	1.040685	0.9646248

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5209404 278.3    8529671 455.6  8529671 455.6
Vcells 9714724  74.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5209380 278.3    8529671 455.6  8529671 455.6
Vcells 9814777  74.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

Table: Benchmarking of colDiffs_X_S(), colDiffs(X, rows, cols)() and colDiffs(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	colDiffs_X_S	0.107511	0.1167435	0.1380288	0.1315225	0.1482610	0.420001
2	colDiffs(X, rows, cols)	0.130151	0.1428920	0.1642985	0.1585660	0.1778200	0.307284
3	colDiffs(X[rows, cols])	0.204619	0.2275180	0.2604196	0.2573770	0.2754015	0.347143

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	colDiffs(X, rows, cols)	1.210583	1.223982	1.190321	1.205619	1.199371	0.7316268
3	colDiffs(X[rows, cols])	1.903238	1.948871	1.886706	1.956905	1.857545	0.8265290

Table: Benchmarking of rowDiffs_X_S(), rowDiffs(X, cols, rows)() and rowDiffs(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	rowDiffs(X, cols, rows)	0.106067	0.1119765	0.1276081	0.1191835	0.1383565	0.286915
1	rowDiffs_X_S	0.126863	0.1413660	0.1584162	0.1575735	0.1701140	0.238276
3	rowDiffs(X[cols, rows])	0.220626	0.2405685	0.2703636	0.2621900	0.2886485	0.407271

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowDiffs_X_S	1.196065	1.262461	1.241428	1.322108	1.229534	0.8304759
3	rowDiffs(X[cols, rows])	2.080063	2.148384	2.118703	2.199885	2.086266	1.4194831

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	107.511	116.7435	138.0288	131.5225	148.261	420.001
2	rowDiffs_X_S	126.863	141.3660	158.4162	157.5735	170.114	238.276

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.00	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rowDiffs_X_S	1.18	1.210911	1.147705	1.198073	1.147396	0.5673225

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 5209614 278.3    8529671 455.6  8529671 455.6
Vcells 9714865  74.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs_X_S = colDiffs(X_S), `colDiffs(X, rows, cols)` = colDiffs(X, 
+     rows = rows, cols = cols), `colDiffs(X[rows, cols])` = colDiffs(X[rows, cols]), unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5209590 278.3    8529671 455.6  8529671 455.6
Vcells 9814918  74.9   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs_X_S = rowDiffs(X_S), `rowDiffs(X, cols, rows)` = rowDiffs(X, 
+     rows = cols, cols = rows), `rowDiffs(X[cols, rows])` = rowDiffs(X[cols, rows]), unit = "ms")

Table: Benchmarking of colDiffs_X_S(), colDiffs(X, rows, cols)() and colDiffs(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	colDiffs_X_S	0.103457	0.1127685	0.1273864	0.1224880	0.1348785	0.194025
2	colDiffs(X, rows, cols)	0.131980	0.1468910	0.1644894	0.1609465	0.1724850	0.273805
3	colDiffs(X[rows, cols])	0.197955	0.2158190	0.2466904	0.2428345	0.2649460	0.433847

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	colDiffs(X, rows, cols)	1.275699	1.302589	1.291263	1.313978	1.278818	1.411184
3	colDiffs(X[rows, cols])	1.913404	1.913823	1.936552	1.982517	1.964331	2.236037

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

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	0.107927	0.1174900	0.1338860	0.1284205	0.1420660	0.321564
1	rowDiffs_X_S	0.134344	0.1399150	0.1629436	0.1591370	0.1757520	0.227270
3	rowDiffs(X[cols, rows])	0.229788	0.2465015	0.2803006	0.2730560	0.3048045	0.382753

	expr	min	lq	mean	median	uq	max
2	rowDiffs(X, cols, rows)	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	rowDiffs_X_S	1.244767	1.190867	1.217033	1.239187	1.237115	0.7067644
3	rowDiffs(X[cols, rows])	2.129106	2.098064	2.093577	2.126265	2.145513	1.1902856

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

Table: Benchmarking of colDiffs_X_S() and rowDiffs_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	colDiffs_X_S	103.457	112.7685	127.3864	122.488	134.8785	194.025
2	rowDiffs_X_S	134.344	139.9150	162.9436	159.137	175.7520	227.270

	expr	min	lq	mean	median	uq	max
1	colDiffs_X_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs_X_S	1.298549	1.240728	1.279129	1.299205	1.303039	1.171344

Figure: Benchmarking of colDiffs_X_S() and rowDiffs_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 21.96 secs.

Reproducibility

To reproduce this report, do:

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

---

Copyright Dongcan Jiang. Last updated on 2021-08-25 18:55:08 (+0200 UTC). Powered by RSP.