matrixStats.benchmarks

---

colDiffs() and rowDiffs() benchmarks

This report benchmark the performance of colDiffs() and rowDiffs() against alternative methods.

Alternative methods

• apply() + diff()
• apply() + diff2()
• diff()

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"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5221589 278.9    8529671 455.6  8529671 455.6
Vcells 9980577  76.2   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5212291 278.4    8529671 455.6  8529671 455.6
Vcells 9950246  76.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.001954	0.0024945	0.0030917	0.0030555	0.0034100	0.011732
4	diff	0.007981	0.0095225	0.0111544	0.0107775	0.0116385	0.055725
3	apply+diff2	0.048816	0.0529325	0.0573205	0.0557615	0.0584190	0.179273
2	apply+diff	0.086678	0.0901780	0.0972666	0.0953810	0.1018200	0.170467

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000
4	diff	4.084442	3.817398	3.607885	3.527246	3.41305	4.749829
3	apply+diff2	24.982600	21.219683	18.540300	18.249550	17.13167	15.280685
2	apply+diff	44.359263	36.150732	31.460851	31.216168	29.85924	14.530089

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() 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	0.001989	0.0027065	0.0031962	0.0032110	0.0035955	0.009475
4	diff + t	0.010558	0.0126440	0.0146196	0.0144855	0.0160235	0.033993
3	apply+diff2	0.049002	0.0529945	0.0570272	0.0562080	0.0587405	0.149900
2	apply+diff	0.087192	0.0923010	0.0993132	0.0983380	0.1041760	0.141377

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff + t	5.308195	4.671716	4.574047	4.511211	4.456543	3.587652
3	apply+diff2	24.636501	19.580454	17.842194	17.504827	16.337227	15.820581
2	apply+diff	43.837104	34.103455	31.072277	30.625350	28.973995	14.921055

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on integer+10x10 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	1.954	2.4945	3.09167	3.0555	3.4100	11.732
2	rowDiffs	1.989	2.7065	3.19620	3.2110	3.5955	9.475

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.00000	1.000000	1.000000	1.0000000
2	rowDiffs	1.017912	1.084987	1.03381	1.050892	1.054399	0.8076202

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

100x100 integer matrix

> X <- data[["100x100"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5210937 278.3    8529671 455.6  8529671 455.6
Vcells 9567136  73.0   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5210913 278.3    8529671 455.6  8529671 455.6
Vcells 9572149  73.1   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.022720	0.0243530	0.0267945	0.0253725	0.0268310	0.043121
4	diff	0.087568	0.0914170	0.1019660	0.0957025	0.1091395	0.180788
3	apply+diff2	0.274411	0.2865855	0.3192940	0.3014605	0.3363070	0.551258
2	apply+diff	0.682734	0.7026075	0.7717211	0.7299490	0.7967990	1.197832

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff	3.854225	3.753829	3.805481	3.771899	4.067664	4.192574
3	apply+diff2	12.077949	11.767975	11.916397	11.881387	12.534270	12.783980
2	apply+diff	30.049912	28.850963	28.801464	28.769298	29.696955	27.778391

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() on integer+100x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowDiffs	0.033468	0.0356325	0.0403932	0.0377810	0.0441705	0.062311
4	diff + t	0.101952	0.1059645	0.1162867	0.1111445	0.1207780	0.180062
3	apply+diff2	0.276969	0.2889045	0.3169044	0.2996115	0.3304000	0.468003
2	apply+diff	0.687461	0.7022585	0.7748586	0.7258275	0.8050630	1.316043

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff + t	3.046253	2.973816	2.878867	2.941809	2.734359	2.889731
3	apply+diff2	8.275636	8.107893	7.845485	7.930216	7.480105	7.510760
2	apply+diff	20.540845	19.708370	19.182887	19.211442	18.226260	21.120557

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on integer+100x100 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	22.720	24.3530	26.79451	25.3725	26.8310	43.121
2	rowDiffs	33.468	35.6325	40.39322	37.7810	44.1705	62.311

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs	1.473063	1.463167	1.507518	1.489053	1.646249	1.445027

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

1000x10 integer matrix

> X <- data[["1000x10"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5211707 278.4    8529671 455.6  8529671 455.6
Vcells 9570927  73.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5211695 278.4    8529671 455.6  8529671 455.6
Vcells 9575960  73.1   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.022243	0.0262515	0.0297688	0.029234	0.033370	0.046911
4	diff	0.091271	0.1035955	0.1154996	0.111918	0.120278	0.206659
3	apply+diff2	0.120975	0.1399330	0.1568516	0.152697	0.163107	0.308744
2	apply+diff	0.253351	0.2852320	0.3318267	0.321153	0.360368	0.947031

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff	4.103358	3.946270	3.879888	3.828351	3.604375	4.405342
3	apply+diff2	5.438790	5.330476	5.268994	5.223267	4.887833	6.581484
2	apply+diff	11.390145	10.865360	11.146799	10.985599	10.799161	20.187824

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() 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	0.034183	0.0372740	0.0416555	0.0403800	0.0450055	0.066722
4	diff + t	0.103999	0.1144365	0.1282346	0.1245210	0.1382080	0.223987
3	apply+diff2	0.122528	0.1343930	0.1513732	0.1456705	0.1608315	0.268266
2	apply+diff	0.260132	0.2768280	0.3205279	0.3149430	0.3478400	0.430764

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff + t	3.042419	3.070143	3.078456	3.083730	3.070914	3.357019
3	apply+diff2	3.584472	3.605543	3.633932	3.607491	3.573597	4.020653
2	apply+diff	7.609982	7.426839	7.694735	7.799480	7.728833	6.456101

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on integer+1000x10 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	22.243	26.2515	29.76879	29.234	33.3700	46.911
2	rowDiffs	34.183	37.2740	41.65548	40.380	45.0055	66.722

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.0000	1.000000	1.000000	1.00000
2	rowDiffs	1.536798	1.419881	1.3993	1.381268	1.348682	1.42231

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

10x1000 integer matrix

> X <- data[["10x1000"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5211961 278.4    8529671 455.6  8529671 455.6
Vcells 9571795  73.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5211937 278.4    8529671 455.6  8529671 455.6
Vcells 9576808  73.1   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.022336	0.0249510	0.0268664	0.0258295	0.0281165	0.047132
4	diff	0.086755	0.0934565	0.0981875	0.0959990	0.0988650	0.164649
3	apply+diff2	1.797929	1.9697590	2.0623319	2.0391190	2.1051500	2.844316
2	apply+diff	4.868596	5.2459445	5.5224501	5.3340625	5.5647270	10.605102

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff	3.884088	3.745601	3.654662	3.716642	3.516263	3.493359
3	apply+diff2	80.494672	78.945092	76.762582	78.945353	74.872406	60.347874
2	apply+diff	217.970810	210.249870	205.552523	206.510482	197.916775	225.008529

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() 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	0.029975	0.0336145	0.0359140	0.0344240	0.0365195	0.067347
4	diff + t	0.097950	0.1095145	0.1172109	0.1133445	0.1190880	0.208593
3	apply+diff2	1.801484	1.9679595	2.1331394	2.0244950	2.0860485	7.038373
2	apply+diff	4.866091	5.2611440	5.4705944	5.3597840	5.5588710	10.491253

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff + t	3.267723	3.257954	3.263657	3.292601	3.260943	3.097287
3	apply+diff2	60.099550	58.544958	59.395795	58.810568	57.121497	104.509080
2	apply+diff	162.338315	156.514123	152.324927	155.699047	152.216514	155.779070

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on integer+10x1000 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	22.336	24.9510	26.86637	25.8295	28.1165	47.132
2	rowDiffs	29.975	33.6145	35.91398	34.4240	36.5195	67.347

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.00000	1.000000	1.000000
2	rowDiffs	1.342004	1.347221	1.336763	1.33274	1.298864	1.428902

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

100x1000 integer matrix

> X <- data[["100x1000"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5212177 278.4    8529671 455.6  8529671 455.6
Vcells 9572361  73.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5212171 278.4    8529671 455.6  8529671 455.6
Vcells 9622404  73.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.220242	0.2393210	0.2456916	0.2451810	0.2506115	0.297203
4	diff	0.837108	0.8952495	1.1099607	0.9266345	0.9504790	18.800757
3	apply+diff2	2.667214	2.8828510	3.1375531	2.9471605	2.9840970	21.870916
2	apply+diff	6.707820	7.2997510	8.0123042	7.4189970	7.5452825	27.194875

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.00000	1.00000	1.00000	1.000000	1.00000
4	diff	3.800855	3.74079	4.51770	3.77939	3.792639	63.25897
3	apply+diff2	12.110379	12.04596	12.77029	12.02035	11.907263	73.58915
2	apply+diff	30.456589	30.50192	32.61123	30.25927	30.107487	91.50269

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() on integer+100x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowDiffs	0.324826	0.350251	0.5336086	0.356782	0.366596	17.692511
4	diff + t	0.986962	1.065435	1.0956329	1.085759	1.114226	1.412372
3	apply+diff2	2.650777	2.889380	3.3135243	2.950258	3.008276	21.220776
2	apply+diff	6.732852	7.320830	7.8556445	7.452088	7.562580	25.741307

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
4	diff + t	3.038433	3.041917	2.053252	3.043201	3.039383	0.0798288
3	apply+diff2	8.160606	8.249455	6.209653	8.269076	8.205971	1.1994214
2	apply+diff	20.727565	20.901666	14.721734	20.886949	20.629194	1.4549267

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on integer+100x1000 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	220.242	239.321	245.6916	245.181	250.6115	297.203
2	rowDiffs	324.826	350.251	533.6086	356.782	366.5960	17692.511

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.00000	1.000000	1.000000	1.000000	1.00000
2	rowDiffs	1.474859	1.46352	2.171864	1.455178	1.462806	59.53006

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

1000x100 integer matrix

> X <- data[["1000x100"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5212401 278.4    8529671 455.6  8529671 455.6
Vcells 9573035  73.1   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5212395 278.4    8529671 455.6  8529671 455.6
Vcells 9623078  73.5   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.210796	0.2235195	0.2327042	0.2320075	0.2400205	0.301566
4	diff	0.811547	0.8648805	0.9887286	0.9038545	0.9344805	8.510391
3	apply+diff2	0.978443	1.0580410	1.1198984	1.1048290	1.1437750	1.412121
2	apply+diff	2.303703	2.4978300	2.8662854	2.5936945	2.7052950	10.602644

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff	3.849916	3.869374	4.248863	3.895799	3.893336	28.220658
3	apply+diff2	4.641658	4.733551	4.812540	4.762040	4.765322	4.682627
2	apply+diff	10.928590	11.174998	12.317289	11.179356	11.271100	35.158619

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() on integer+1000x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowDiffs	0.325121	0.346176	0.4463295	0.3646085	0.3858825	7.986657
4	diff + t	0.934064	1.009519	1.0746035	1.0615935	1.0999650	1.494019
3	apply+diff2	0.985068	1.077364	1.2894496	1.1212320	1.1768665	8.874917
2	apply+diff	2.310003	2.495784	2.8190690	2.5909345	2.7039315	11.189777

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
4	diff + t	2.872973	2.916203	2.407646	2.911598	2.850518	0.1870644
3	apply+diff2	3.029850	3.112186	2.889008	3.075167	3.049805	1.1112180
2	apply+diff	7.105056	7.209581	6.316116	7.106073	7.007137	1.4010589

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on integer+1000x100 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	210.796	223.5195	232.7042	232.0075	240.0205	301.566
2	rowDiffs	325.121	346.1760	446.3295	364.6085	385.8825	7986.657

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
2	rowDiffs	1.542349	1.548751	1.918012	1.571538	1.607706	26.48394

Figure: Benchmarking of colDiffs() and rowDiffs() 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"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5212651 278.4    8529671 455.6  8529671 455.6
Vcells 9689681  74.0   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5212618 278.4    8529671 455.6  8529671 455.6
Vcells 9689779  74.0   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.002080	0.0026625	0.0033289	0.0032840	0.0036695	0.012743
4	diff	0.008685	0.0099565	0.0115552	0.0113600	0.0122175	0.050040
3	apply+diff2	0.051650	0.0547855	0.0577088	0.0566275	0.0600265	0.095607
2	apply+diff	0.090211	0.0942510	0.1012959	0.0981310	0.1063135	0.179519

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
4	diff	4.175481	3.73953	3.471193	3.459196	3.329473	3.926862
3	apply+diff2	24.831731	20.57671	17.335859	17.243453	16.358223	7.502707
2	apply+diff	43.370673	35.39944	30.429524	29.881547	28.972203	14.087656

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() 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	0.001881	0.0026795	0.0031912	0.0032365	0.0035175	0.009450
4	diff + t	0.010592	0.0129275	0.0146411	0.0145545	0.0160270	0.033850
3	apply+diff2	0.049248	0.0521735	0.0564470	0.0550390	0.0579845	0.139601
2	apply+diff	0.087028	0.0912560	0.0982636	0.0976895	0.1026715	0.134803

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff + t	5.631047	4.824594	4.587989	4.496988	4.556361	3.582011
3	apply+diff2	26.181818	19.471357	17.688432	17.005716	16.484577	14.772593
2	apply+diff	46.266879	34.057100	30.792255	30.183686	29.188770	14.264868

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on double+10x10 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	rowDiffs	1.881	2.6795	3.19118	3.2365	3.5175	9.450
1	colDiffs	2.080	2.6625	3.32887	3.2840	3.6695	12.743

	expr	min	lq	mean	median	uq	max
2	rowDiffs	1.000000	1.0000000	1.000000	1.000000	1.000000	1.000000
1	colDiffs	1.105795	0.9936555	1.043147	1.014676	1.043213	1.348466

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

100x100 double matrix

> X <- data[["100x100"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5212870 278.4    8529671 455.6  8529671 455.6
Vcells 9689820  74.0   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5212846 278.4    8529671 455.6  8529671 455.6
Vcells 9699833  74.1   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.023100	0.0249275	0.0273628	0.0259930	0.0283365	0.045033
4	diff	0.051161	0.0545710	0.0622083	0.0577905	0.0668045	0.128612
3	apply+diff2	0.276821	0.2903645	0.3266575	0.3081505	0.3526620	0.574421
2	apply+diff	0.627511	0.6541060	0.7320162	0.6972865	0.7858250	1.221230

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.00000	1.00000	1.000000	1.00000
4	diff	2.214762	2.189189	2.27346	2.22331	2.357542	2.85595
3	apply+diff2	11.983593	11.648360	11.93800	11.85513	12.445503	12.75556
2	apply+diff	27.164978	26.240337	26.75221	26.82593	27.731901	27.11856

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() 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	rowDiffs	0.028255	0.0307465	0.0346609	0.0333050	0.0375870	0.054360
4	diff + t	0.068153	0.0724020	0.0798607	0.0758350	0.0845270	0.125395
3	apply+diff2	0.285219	0.2982425	0.3285841	0.3114945	0.3449940	0.491718
2	apply+diff	0.645122	0.6712770	0.7397274	0.6928275	0.7856995	1.297721

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff + t	2.412069	2.354805	2.304057	2.276985	2.248836	2.306751
3	apply+diff2	10.094461	9.700047	9.479962	9.352785	9.178546	9.045585
2	apply+diff	22.832136	21.832631	21.341833	20.802507	20.903491	23.872719

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on double+100x100 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	23.100	24.9275	27.36283	25.993	28.3365	45.033
2	rowDiffs	28.255	30.7465	34.66091	33.305	37.5870	54.360

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.00000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs	1.22316	1.233437	1.266715	1.281306	1.326452	1.207115

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

1000x10 double matrix

> X <- data[["1000x10"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5213076 278.5    8529671 455.6  8529671 455.6
Vcells 9690852  74.0   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5213070 278.5    8529671 455.6  8529671 455.6
Vcells 9700895  74.1   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.023613	0.0268990	0.0302743	0.0293400	0.034039	0.042942
4	diff	0.055417	0.0600370	0.0683621	0.0666035	0.071845	0.137372
3	apply+diff2	0.131302	0.1448125	0.1641396	0.1611825	0.171947	0.287777
2	apply+diff	0.206308	0.2217895	0.2566725	0.2560955	0.278098	0.360161

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff	2.346885	2.231942	2.258087	2.270058	2.110667	3.199013
3	apply+diff2	5.560581	5.383564	5.421738	5.493609	5.051470	6.701528
2	apply+diff	8.737052	8.245269	8.478219	8.728545	8.169982	8.387150

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() 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	rowDiffs	0.030439	0.0333965	0.0370461	0.0362580	0.0397695	0.057650
4	diff + t	0.070721	0.0791585	0.0890181	0.0856025	0.0956845	0.184646
3	apply+diff2	0.128627	0.1424815	0.1609605	0.1564495	0.1709555	0.269732
2	apply+diff	0.203164	0.2225625	0.2577868	0.2579720	0.2822365	0.355501

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff + t	2.323368	2.370263	2.402903	2.360927	2.405977	3.202879
3	apply+diff2	4.225730	4.266360	4.344873	4.314896	4.298659	4.678786
2	apply+diff	6.674464	6.664246	6.958543	7.114899	7.096808	6.166539

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on double+1000x10 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	23.613	26.8990	30.27434	29.340	34.0390	42.942
2	rowDiffs	30.439	33.3965	37.04608	36.258	39.7695	57.650

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs	1.289078	1.241552	1.223679	1.235787	1.168351	1.342509

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

10x1000 double matrix

> X <- data[["10x1000"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5213316 278.5    8529671 455.6  8529671 455.6
Vcells 9692122  74.0   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5213292 278.5    8529671 455.6  8529671 455.6
Vcells 9702135  74.1   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.024531	0.0267080	0.0294914	0.0283005	0.0312220	0.056440
4	diff	0.054032	0.0588175	0.0640694	0.0623700	0.0657035	0.117145
3	apply+diff2	1.821629	2.0057890	2.0758672	2.0595995	2.1228035	2.777601
2	apply+diff	4.818522	5.2300530	5.5687312	5.4151435	5.6377065	10.713772

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff	2.202601	2.202243	2.172475	2.203848	2.104398	2.075567
3	apply+diff2	74.258245	75.100681	70.388853	72.776082	67.990632	49.213342
2	apply+diff	196.425828	195.823461	188.825468	191.344446	180.568397	189.825868

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() 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	rowDiffs	0.027112	0.0289850	0.0313485	0.0299175	0.0324550	0.058580
4	diff + t	0.071328	0.0781565	0.0849833	0.0835660	0.0877535	0.125718
3	apply+diff2	1.803224	1.9961200	2.1511537	2.0514605	2.1154205	7.177878
2	apply+diff	4.844920	5.2544145	5.4940539	5.4185500	5.6013575	10.422620

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.00000	1.000000	1.000000	1.000000
4	diff + t	2.630865	2.696446	2.71092	2.793215	2.703851	2.146091
3	apply+diff2	66.510180	68.867345	68.62060	68.570586	65.180111	122.531205
2	apply+diff	178.700207	181.280473	175.25726	181.116403	172.588430	177.921133

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on double+10x1000 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	24.531	26.708	29.49142	28.3005	31.222	56.44
2	rowDiffs	27.112	28.985	31.34851	29.9175	32.455	58.58

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs	1.105214	1.085255	1.062971	1.057137	1.039491	1.037916

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

100x1000 double matrix

> X <- data[["100x1000"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5213528 278.5    8529671 455.6  8529671 455.6
Vcells 9692247  74.0   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5213522 278.5    8529671 455.6  8529671 455.6
Vcells 9792290  74.8   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.215833	0.2410510	0.4041848	0.2538330	0.2675530	15.18175
4	diff	0.489540	0.5349305	0.7167858	0.5606355	0.5856455	15.57865
3	apply+diff2	2.646576	2.8985810	3.2437154	3.0101815	3.1981280	19.16186
2	apply+diff	6.129846	6.7155305	7.8910013	6.9547935	7.3323635	24.16777

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff	2.268143	2.219159	1.773411	2.208678	2.188895	1.026143
3	apply+diff2	12.262147	12.024762	8.025328	11.858905	11.953250	1.262164
2	apply+diff	28.400875	27.859376	19.523253	27.399091	27.405275	1.591896

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() on double+100x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowDiffs	0.265864	0.291026	0.3117314	0.3055070	0.3185970	0.412818
4	diff + t	0.636402	0.712100	0.7476812	0.7408175	0.7808475	1.034439
3	apply+diff2	2.745435	2.978209	3.5760799	3.0764305	3.3085420	18.899369
2	apply+diff	6.243330	6.754134	7.8876042	7.0142665	7.2528700	23.494539

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
4	diff + t	2.393713	2.44686	2.398479	2.424879	2.450894	2.505799
3	apply+diff2	10.326464	10.23348	11.471671	10.069918	10.384724	45.781359
2	apply+diff	23.483172	23.20801	25.302567	22.959430	22.765029	56.912584

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on double+100x1000 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	215.833	241.051	404.1848	253.833	267.553	15181.753
2	rowDiffs	265.864	291.026	311.7314	305.507	318.597	412.818

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	rowDiffs	1.231804	1.207321	0.7712596	1.203575	1.190781	0.0271917

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

1000x100 double matrix

> X <- data[["1000x100"]]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5213764 278.5    8529671 455.6  8529671 455.6
Vcells 9693731  74.0   31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colDiffs = colDiffs(X), `apply+diff` = apply(X, MARGIN = 2L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 2L, FUN = diff2), diff = diff(X), unit = "ms")
> X <- t(X)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 5213746 278.5    8529671 455.6  8529671 455.6
Vcells 9793754  74.8   31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowDiffs = rowDiffs(X), `apply+diff` = apply(X, MARGIN = 1L, FUN = diff), 
+     `apply+diff2` = apply(X, MARGIN = 1L, FUN = diff2), `diff + t` = diff(t(X)), unit = "ms")

Table: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() 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	0.219531	0.2375990	0.2522123	0.2491625	0.2655255	0.353970
4	diff	0.478195	0.5247625	0.6945688	0.5581735	0.5891675	7.334720
3	apply+diff2	1.076347	1.1491805	1.2472880	1.2362115	1.3243010	1.573288
2	apply+diff	1.827761	1.9305060	2.5260755	2.1793120	2.3262090	8.918474

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	diff	2.178257	2.208606	2.753905	2.240199	2.218873	20.721304
3	apply+diff2	4.902939	4.836639	4.945389	4.961467	4.987472	4.444693
2	apply+diff	8.325753	8.125059	10.015671	8.746549	8.760774	25.195565

Table: Benchmarking of rowDiffs(), apply+diff(), apply+diff2() and diff + t() 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	rowDiffs	0.272925	0.2934385	0.3149624	0.3074485	0.3376200	0.401050
4	diff + t	0.611892	0.6760500	0.8569438	0.7228820	0.7589485	7.506957
3	apply+diff2	1.085350	1.1983680	1.4240821	1.2975220	1.3898940	7.975555
2	apply+diff	1.823827	1.9672945	2.4480114	2.1906855	2.3622570	9.479181

	expr	min	lq	mean	median	uq	max
1	rowDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.00000
4	diff + t	2.241979	2.303890	2.720781	2.351230	2.247937	18.71826
3	apply+diff2	3.976733	4.083881	4.521435	4.220291	4.116741	19.88668
2	apply+diff	6.682521	6.704282	7.772393	7.125374	6.996792	23.63591

Figure: Benchmarking of colDiffs(), apply+diff(), apply+diff2() and diff() on double+1000x100 data as well as rowDiffs(), apply+diff(), apply+diff2() and diff + t() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colDiffs() and rowDiffs() 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	219.531	237.5990	252.2123	249.1625	265.5255	353.97
2	rowDiffs	272.925	293.4385	314.9624	307.4485	337.6200	401.05

	expr	min	lq	mean	median	uq	max
1	colDiffs	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowDiffs	1.243219	1.235016	1.248799	1.233928	1.271516	1.133006

Figure: Benchmarking of colDiffs() and rowDiffs() 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 32.79 secs.

Reproducibility

To reproduce this report, do:

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

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 18:55:42 (+0200 UTC). Powered by RSP.