matrixStats.benchmarks

---

colSums2() and rowSums2() benchmarks

This report benchmark the performance of colSums2() and rowSums2() against alternative methods.

Alternative methods

• apply() + sum()
• colSums() and rowSums()
• .colSums() and .rowSums()

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  5320133 284.2    7916910 422.9  7916910 422.9
Vcells 10632655  81.2   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5306443 283.4    7916910 422.9  7916910 422.9
Vcells 10587500  80.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+10x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2	0.002443	0.0028260	0.0034775	0.0034740	0.0038005	0.014816
2	.colSums	0.002986	0.0033955	0.0039280	0.0038030	0.0042120	0.019052
3	colSums	0.005357	0.0059335	0.0066121	0.0063735	0.0071510	0.016620
4	apply+sum	0.035054	0.0367040	0.0384957	0.0374870	0.0381930	0.112776

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.colSums	1.222268	1.201522	1.129534	1.094704	1.108275	1.285907
3	colSums	2.192796	2.099611	1.901359	1.834629	1.881595	1.121760
4	apply+sum	14.348752	12.987969	11.069770	10.790731	10.049467	7.611771

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() on integer+10x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2	0.002369	0.002815	0.0034117	0.0033630	0.0037175	0.014173
2	.rowSums	0.004153	0.004480	0.0048022	0.0046295	0.0048510	0.016410
3	rowSums	0.006330	0.006822	0.0077290	0.0073640	0.0076690	0.047929
4	apply+sum	0.033702	0.035261	0.0374024	0.0360015	0.0373740	0.096818

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.753060	1.591474	1.407587	1.376598	1.304909	1.157835
3	rowSums	2.672013	2.423446	2.265451	2.189712	2.062946	3.381712
4	apply+sum	14.226256	12.526110	10.963103	10.705174	10.053531	6.831158

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+10x10 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() on integer+10x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	rowSums2	2.369	2.815	3.41166	3.363	3.7175	14.173
1	colSums2	2.443	2.826	3.47755	3.474	3.8005	14.816

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colSums2	1.031237	1.003908	1.019313	1.033006	1.022327	1.045368

Figure: Benchmarking of colSums2() and rowSums2() 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  5305026 283.4    7916910 422.9  7916910 422.9
Vcells 10204316  77.9   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5305020 283.4    7916910 422.9  7916910 422.9
Vcells 10209359  77.9   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	.colSums	0.014013	0.0169955	0.0230925	0.0200490	0.0296605	0.038990
3	colSums	0.016865	0.0200345	0.0282121	0.0246215	0.0355295	0.056384
1	colSums2	0.015264	0.0194205	0.0253197	0.0258200	0.0297810	0.042907
4	apply+sum	0.211092	0.2498680	0.3446062	0.3099720	0.4590165	0.649330

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.203525	1.178812	1.221701	1.228066	1.197873	1.446114
1	colSums2	1.089274	1.142685	1.096448	1.287845	1.004063	1.100462
4	apply+sum	15.064012	14.702009	14.922881	15.460721	15.475683	16.653757

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	rowSums2	0.013112	0.0146740	0.0157031	0.0154150	0.016288	0.029076
2	.rowSums	0.037140	0.0385595	0.0418892	0.0408705	0.043711	0.061806
3	rowSums	0.039234	0.0417435	0.0442904	0.0442865	0.045929	0.056231
4	apply+sum	0.174624	0.1864420	0.1999740	0.1976550	0.207283	0.315519

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	2.832520	2.627743	2.667569	2.651346	2.683632	2.125671
3	rowSums	2.992221	2.844725	2.820482	2.872948	2.819806	1.933932
4	apply+sum	13.317877	12.705602	12.734665	12.822251	12.726117	10.851527

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+100x100 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() on integer+100x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	rowSums2	13.112	14.6740	15.70312	15.415	16.288	29.076
1	colSums2	15.264	19.4205	25.31970	25.820	29.781	42.907

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	colSums2	1.164125	1.323463	1.612399	1.674992	1.828401	1.475684

Figure: Benchmarking of colSums2() and rowSums2() 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  5304457 283.3    7916910 422.9  7916910 422.9
Vcells 10182305  77.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5304445 283.3    7916910 422.9  7916910 422.9
Vcells 10187338  77.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.012793	0.0138765	0.0151997	0.0147500	0.0155460	0.023290
3	colSums	0.015114	0.0165615	0.0185141	0.0174700	0.0187055	0.050554
1	colSums2	0.016597	0.0181430	0.0194547	0.0191415	0.0197975	0.036493
4	apply+sum	0.102830	0.1107745	0.1190742	0.1142830	0.1222910	0.302068

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.181427	1.193493	1.218055	1.184407	1.203236	2.170631
1	colSums2	1.297350	1.307462	1.279936	1.297729	1.273479	1.566896
4	apply+sum	8.037990	7.982885	7.833978	7.748000	7.866396	12.969858

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() on integer+1000x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2	0.013856	0.0157020	0.0177937	0.0171130	0.0184795	0.047828
4	apply+sum	0.085355	0.0956420	0.1044608	0.1020715	0.1101325	0.244464
2	.rowSums	0.139088	0.1514185	0.1666585	0.1659255	0.1775225	0.219063
3	rowSums	0.141147	0.1564940	0.1697981	0.1685615	0.1797755	0.241362

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	apply+sum	6.160147	6.091071	5.870658	5.964559	5.959712	5.111315
2	.rowSums	10.038106	9.643262	9.366148	9.695875	9.606456	4.580225
3	rowSums	10.186706	9.966501	9.542592	9.849909	9.728375	5.046458

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+1000x10 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() on integer+1000x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	rowSums2	13.856	15.702	17.79371	17.1130	18.4795	47.828
1	colSums2	16.597	18.143	19.45466	19.1415	19.7975	36.493

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.00000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	colSums2	1.19782	1.155458	1.093345	1.118536	1.071322	0.7630049

Figure: Benchmarking of colSums2() and rowSums2() 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  5304674 283.4    7916910 422.9  7916910 422.9
Vcells 10183108  77.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5304668 283.3    7916910 422.9  7916910 422.9
Vcells 10188151  77.8   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.011768	0.0122945	0.0140559	0.0132670	0.0152755	0.022618
1	colSums2	0.011852	0.0125070	0.0149309	0.0141490	0.0163095	0.028170
3	colSums	0.013389	0.0144525	0.0174395	0.0167285	0.0189630	0.048868
4	apply+sum	0.901248	0.9727045	1.0376723	0.9987180	1.0757920	1.455643

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.00000	1.000000
1	colSums2	1.007138	1.017284	1.062247	1.066481	1.06769	1.245468
3	colSums	1.137746	1.175526	1.240723	1.260911	1.24140	2.160580
4	apply+sum	76.584636	79.117044	73.824522	75.278360	70.42598	64.357724

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() on integer+10x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2	0.013819	0.0153655	0.0208467	0.0177800	0.0240845	0.069286
2	.rowSums	0.028310	0.0298670	0.0333621	0.0317125	0.0349430	0.056816
3	rowSums	0.030562	0.0324720	0.0395548	0.0384650	0.0434175	0.085054
4	apply+sum	1.056236	1.1222810	1.2903119	1.1987045	1.3752205	2.546222

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.rowSums	2.048629	1.943770	1.600354	1.783605	1.450850	0.8200214
3	rowSums	2.211593	2.113306	1.897412	2.163386	1.802715	1.2275784
4	apply+sum	76.433606	73.039016	61.895289	67.418701	57.099815	36.7494443

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+10x1000 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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	colSums2	11.852	12.5070	14.93087	14.149	16.3095	28.170
2	rowSums2	13.819	15.3655	20.84669	17.780	24.0845	69.286

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rowSums2	1.165964	1.228552	1.396214	1.256626	1.476716	2.459567

Figure: Benchmarking of colSums2() and rowSums2() 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  5304916 283.4    7916910 422.9  7916910 422.9
Vcells 10183721  77.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5304898 283.4    7916910 422.9  7916910 422.9
Vcells 10233744  78.1   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	.colSums	0.087179	0.0878230	0.0960942	0.0887540	0.0985205	0.170537
3	colSums	0.089127	0.0901555	0.0986730	0.0929335	0.0999985	0.170725
1	colSums2	0.099686	0.1003240	0.1087257	0.1021290	0.1117535	0.166117
4	apply+sum	1.401935	1.4424960	1.5671584	1.4758180	1.5615780	2.647030

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	colSums	1.022345	1.026559	1.026836	1.047091	1.015002	1.0011024
1	colSums2	1.143463	1.142343	1.131449	1.150697	1.134317	0.9740819
4	apply+sum	16.081109	16.425037	16.308566	16.628186	15.850285	15.5217343

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	rowSums2	0.105325	0.1066975	0.1218370	0.1099685	0.1277735	0.264190
2	.rowSums	0.219689	0.2202540	0.2390576	0.2221425	0.2397650	0.419237
3	rowSums	0.221649	0.2229495	0.2433800	0.2269330	0.2475130	0.425284
4	apply+sum	1.416842	1.4426810	1.5820874	1.4741565	1.6123855	3.018600

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	2.085820	2.064284	1.962110	2.020056	1.876485	1.586877
3	rowSums	2.104429	2.089548	1.997587	2.063618	1.937123	1.609766
4	apply+sum	13.452096	13.521226	12.985280	13.405261	12.619092	11.425868

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+100x1000 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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	colSums2	99.686	100.3240	108.7257	102.1290	111.7535	166.117
2	rowSums2	105.325	106.6975	121.8370	109.9685	127.7735	264.190

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.00000	1.000000	1.000000	1.000000
2	rowSums2	1.056568	1.063529	1.12059	1.076761	1.143351	1.590385

Figure: Benchmarking of colSums2() and rowSums2() 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  5305140 283.4    7916910 422.9  7916910 422.9
Vcells 10184418  77.8   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5305122 283.4    7916910 422.9  7916910 422.9
Vcells 10234441  78.1   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	.colSums	0.080825	0.0814200	0.0925807	0.0858115	0.1015380	0.158800
3	colSums	0.083017	0.0845695	0.0990918	0.0927420	0.1084165	0.166100
1	colSums2	0.105335	0.1067730	0.1236165	0.1175175	0.1360960	0.224615
4	apply+sum	0.626591	0.6417680	0.7705930	0.7209185	0.8352540	2.095169

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.027120	1.038682	1.070329	1.080764	1.067743	1.045970
1	colSums2	1.303248	1.311385	1.335230	1.369484	1.340345	1.414452
4	apply+sum	7.752440	7.882191	8.323474	8.401188	8.226024	13.193759

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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	rowSums2	0.109734	0.1125235	0.1265850	0.1165155	0.1328770	0.225612
2	.rowSums	0.320026	0.3208555	0.3533995	0.3292035	0.3697975	0.538800
3	rowSums	0.322186	0.3234910	0.3569596	0.3311780	0.3898485	0.497132
4	apply+sum	0.628772	0.6369455	0.7241115	0.6439215	0.7930500	1.346452

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	2.916380	2.851453	2.791795	2.825405	2.783006	2.388171
3	rowSums	2.936064	2.874875	2.819919	2.842351	2.933905	2.203482
4	apply+sum	5.729965	5.660555	5.720356	5.526488	5.968301	5.967998

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on integer+1000x100 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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
2	rowSums2	109.734	112.5235	126.5850	116.5155	132.877	225.612
1	colSums2	105.335	106.7730	123.6165	117.5175	136.096	224.615

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.0000000	1.0000000	1.0000000	1.0000	1.000000	1.0000000
1	colSums2	0.9599122	0.9488951	0.9765488	1.0086	1.024225	0.9955809

Figure: Benchmarking of colSums2() and rowSums2() 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  5305359 283.4    7916910 422.9  7916910 422.9
Vcells 10300314  78.6   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5305344 283.4    7916910 422.9  7916910 422.9
Vcells 10300442  78.6   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+10x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2	0.002527	0.0029605	0.0036175	0.0036345	0.0039185	0.014904
2	.colSums	0.002988	0.0033955	0.0040478	0.0036795	0.0041665	0.020700
3	colSums	0.005400	0.0059695	0.0068039	0.0065435	0.0073040	0.016606
4	apply+sum	0.036608	0.0373395	0.0395886	0.0377800	0.0384100	0.102978

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.colSums	1.182430	1.146935	1.118948	1.012381	1.063289	1.388889
3	colSums	2.136921	2.016382	1.880837	1.800385	1.863979	1.114197
4	apply+sum	14.486743	12.612565	10.943731	10.394827	9.802220	6.909420

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() 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
2	.rowSums	0.002414	0.0027580	0.0030384	0.0029855	0.0031470	0.008872
1	rowSums2	0.002421	0.0027590	0.0034438	0.0034360	0.0037070	0.015525
3	rowSums	0.004601	0.0050715	0.0059537	0.0056310	0.0060570	0.032032
4	apply+sum	0.033402	0.0359200	0.0378522	0.0365800	0.0372745	0.099233

	expr	min	lq	mean	median	uq	max
2	.rowSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	rowSums2	1.002900	1.000363	1.133431	1.150896	1.177947	1.749887
3	rowSums	1.905965	1.838833	1.959521	1.886116	1.924690	3.610460
4	apply+sum	13.836785	13.023930	12.458157	12.252554	11.844455	11.184964

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+10x10 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() 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	rowSums2	2.421	2.7590	3.44376	3.4360	3.7070	15.525
1	colSums2	2.527	2.9605	3.61747	3.6345	3.9185	14.904

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.00
1	colSums2	1.043784	1.073034	1.050442	1.057771	1.057054	0.96

Figure: Benchmarking of colSums2() and rowSums2() 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  5305578 283.4    7916910 422.9  7916910 422.9
Vcells 10301331  78.6   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5305572 283.4    7916910 422.9  7916910 422.9
Vcells 10311374  78.7   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	.colSums	0.010424	0.011555	0.0123870	0.0120555	0.0131500	0.018610
3	colSums	0.013015	0.014233	0.0155818	0.0150615	0.0163210	0.035795
1	colSums2	0.014755	0.015874	0.0172287	0.0166125	0.0178500	0.031936
4	apply+sum	0.205700	0.216111	0.2353752	0.2211880	0.2480635	0.366932

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.248561	1.231761	1.257912	1.249347	1.241141	1.923428
1	colSums2	1.415483	1.373778	1.390869	1.378002	1.357414	1.716067
4	apply+sum	19.733308	18.702813	19.001796	18.347476	18.864144	19.716926

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() on double+100x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2	0.014657	0.0158545	0.0171775	0.0166780	0.0178195	0.034507
2	.rowSums	0.027967	0.0292890	0.0315811	0.0302465	0.0320240	0.060768
3	rowSums	0.030290	0.0320290	0.0338461	0.0327030	0.0346775	0.051010
4	apply+sum	0.198443	0.2073905	0.2251430	0.2162170	0.2309470	0.377271

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.908098	1.847362	1.838513	1.813557	1.797132	1.761034
3	rowSums	2.066589	2.020183	1.970370	1.960847	1.946042	1.478251
4	apply+sum	13.539128	13.080860	13.106837	12.964204	12.960352	10.933173

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+100x100 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() on double+100x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2	14.755	15.8740	17.22869	16.6125	17.8500	31.936
2	rowSums2	14.657	15.8545	17.17752	16.6780	17.8195	34.507

	expr	min	lq	mean	median	uq	max
1	colSums2	1.0000000	1.0000000	1.00000	1.000000	1.0000000	1.000000
2	rowSums2	0.9933582	0.9987716	0.99703	1.003943	0.9982913	1.080505

Figure: Benchmarking of colSums2() and rowSums2() 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  5305801 283.4    7916910 422.9  7916910 422.9
Vcells 10302434  78.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5305795 283.4    7916910 422.9  7916910 422.9
Vcells 10312477  78.7   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.012391	0.0134465	0.0147897	0.0142520	0.0153970	0.029268
3	colSums	0.014995	0.0161215	0.0177038	0.0171055	0.0182125	0.036283
1	colSums2	0.016480	0.0181825	0.0194824	0.0192875	0.0201935	0.033885
4	apply+sum	0.108932	0.1141135	0.1239832	0.1232055	0.1294040	0.215156

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.210152	1.198937	1.197031	1.200217	1.182860	1.239682
1	colSums2	1.329998	1.352211	1.317296	1.353319	1.311522	1.157749
4	apply+sum	8.791219	8.486484	8.383067	8.644787	8.404494	7.351237

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() on double+1000x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2	0.016584	0.0179705	0.0197192	0.0193890	0.0205360	0.036673
2	.rowSums	0.030400	0.0317655	0.0349252	0.0345290	0.0367375	0.051910
3	rowSums	0.032999	0.0346235	0.0379493	0.0371175	0.0393060	0.056746
4	apply+sum	0.108867	0.1159555	0.1274138	0.1274865	0.1327665	0.237947

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.833092	1.767647	1.771130	1.780855	1.788932	1.415483
3	rowSums	1.989810	1.926685	1.924489	1.914359	1.914005	1.547351
4	apply+sum	6.564580	6.452547	6.461422	6.575197	6.465061	6.488343

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+1000x10 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() on double+1000x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2	16.480	18.1825	19.48244	19.2875	20.1935	33.885
2	rowSums2	16.584	17.9705	19.71916	19.3890	20.5360	36.673

	expr	min	lq	mean	median	uq	max
1	colSums2	1.000000	1.0000000	1.00000	1.000000	1.000000	1.000000
2	rowSums2	1.006311	0.9883404	1.01215	1.005262	1.016961	1.082278

Figure: Benchmarking of colSums2() and rowSums2() 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  5306024 283.4    7916910 422.9  7916910 422.9
Vcells 10302582  78.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5306018 283.4    7916910 422.9  7916910 422.9
Vcells 10312625  78.7   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() 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
2	.colSums	0.008640	0.0092280	0.0111564	0.0101145	0.0123455	0.021484
3	colSums	0.010223	0.0114865	0.0150963	0.0128015	0.0157860	0.054324
1	colSums2	0.011405	0.0124235	0.0155689	0.0134150	0.0172405	0.037127
4	apply+sum	0.949944	0.9947425	1.1029722	1.0220600	1.1756230	1.730377

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.183218	1.244744	1.353160	1.265658	1.278685	2.528579
1	colSums2	1.320023	1.346283	1.395523	1.326314	1.396501	1.728123
4	apply+sum	109.947222	107.796110	98.864965	101.048989	95.226844	80.542590

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() on double+10x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2	0.011325	0.0133810	0.0201184	0.0156735	0.0264440	0.063460
2	.rowSums	0.023211	0.0240520	0.0307085	0.0283260	0.0346465	0.056080
3	rowSums	0.025084	0.0267470	0.0357664	0.0299540	0.0415890	0.072642
4	apply+sum	0.928487	0.9807665	1.3581019	1.1250920	1.6066060	2.879927

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	.rowSums	2.049536	1.797474	1.526388	1.807254	1.310184	0.8837063
3	rowSums	2.214923	1.998879	1.777792	1.911124	1.572720	1.1446896
4	apply+sum	81.985607	73.295456	67.505430	71.783073	60.755029	45.3817680

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+10x1000 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() on double+10x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colSums2	11.405	12.4235	15.56894	13.4150	17.2405	37.127
2	rowSums2	11.325	13.3810	20.11841	15.6735	26.4440	63.460

	expr	min	lq	mean	median	uq	max
1	colSums2	1.0000000	1.000000	1.000000	1.000000	1.00000	1.000000
2	rowSums2	0.9929855	1.077072	1.292215	1.168356	1.53383	1.709268

Figure: Benchmarking of colSums2() and rowSums2() 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  5306266 283.4    7916910 422.9  7916910 422.9
Vcells 10303906  78.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5306248 283.4    7916910 422.9  7916910 422.9
Vcells 10403929  79.4   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	.colSums	0.063519	0.0644615	0.0745041	0.0693145	0.0805915	0.151046
3	colSums	0.065544	0.0667460	0.0774643	0.0728300	0.0854095	0.113429
1	colSums2	0.099850	0.1019345	0.1142548	0.1085675	0.1203610	0.163218
4	apply+sum	1.444777	1.4960535	1.8489928	1.5252220	1.8728550	19.276073

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	colSums	1.031880	1.035440	1.039732	1.050718	1.059783	0.7509567
1	colSums2	1.571971	1.581324	1.533537	1.566303	1.493470	1.0805847
4	apply+sum	22.745588	23.208481	24.817333	22.004371	23.238865	127.6172358

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() on double+100x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2	0.105869	0.1065635	0.1179462	0.108301	0.1223070	0.222845
2	.rowSums	0.209276	0.2096720	0.2321831	0.210591	0.2481645	0.380039
3	rowSums	0.211461	0.2120145	0.2355641	0.214071	0.2663170	0.341536
4	apply+sum	1.473562	1.4963040	1.8532703	1.510482	1.8528630	19.268951

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	1.976745	1.967578	1.968550	1.944497	2.029029	1.705396
3	rowSums	1.997384	1.989560	1.997216	1.976630	2.177447	1.532617
4	apply+sum	13.918730	14.041431	15.712838	13.947069	15.149280	86.467953

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+100x1000 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() on double+100x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	rowSums2	105.869	106.5635	117.9463	108.3010	122.307	222.845
1	colSums2	99.850	101.9345	114.2548	108.5675	120.361	163.218

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.0000000	1.0000000	1.0000000	1.000000	1.0000000	1.0000000
1	colSums2	0.9431467	0.9565611	0.9687019	1.002461	0.9840892	0.7324284

Figure: Benchmarking of colSums2() and rowSums2() 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  5306478 283.4    7916910 422.9  7916910 422.9
Vcells 10304035  78.7   33191153 253.3 53339345 407.0
> colStats <- microbenchmark(colSums2 = colSums2(X, na.rm = FALSE), .colSums = .colSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), colSums = colSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 2L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
           used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells  5306472 283.4    7916910 422.9  7916910 422.9
Vcells 10404078  79.4   33191153 253.3 53339345 407.0
> rowStats <- microbenchmark(rowSums2 = rowSums2(X, na.rm = FALSE), .rowSums = .rowSums(X, m = nrow(X), 
+     n = ncol(X), na.rm = FALSE), rowSums = rowSums(X, na.rm = FALSE), `apply+sum` = apply(X, MARGIN = 1L, 
+     FUN = sum, na.rm = FALSE), unit = "ms")

Table: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	.colSums	0.077809	0.0796295	0.0931317	0.0855230	0.1012225	0.150511
3	colSums	0.079901	0.0838235	0.0989707	0.0979645	0.1076140	0.144898
1	colSums2	0.105267	0.1110315	0.1277037	0.1263600	0.1387345	0.186246
4	apply+sum	0.684045	0.7195620	0.9312435	0.8170795	0.9337255	9.103571

	expr	min	lq	mean	median	uq	max
2	.colSums	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	colSums	1.026886	1.052669	1.062695	1.145476	1.063143	0.962707
1	colSums2	1.352890	1.394351	1.371216	1.477497	1.370589	1.237424
4	apply+sum	8.791335	9.036375	9.999206	9.553915	9.224486	60.484423

Table: Benchmarking of rowSums2(), .rowSums(), rowSums() and apply+sum() on double+1000x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowSums2	0.107074	0.1089395	0.1279414	0.1202215	0.1370145	0.227836
2	.rowSums	0.214771	0.2154545	0.2465872	0.2255710	0.2708050	0.379276
3	rowSums	0.216914	0.2177035	0.2543414	0.2428245	0.2832430	0.364674
4	apply+sum	0.697127	0.7052320	0.9061603	0.7305075	0.9411070	8.322270

	expr	min	lq	mean	median	uq	max
1	rowSums2	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	.rowSums	2.005818	1.977745	1.927345	1.876295	1.976470	1.664689
3	rowSums	2.025833	1.998389	1.987953	2.019809	2.067248	1.600599
4	apply+sum	6.510703	6.473611	7.082622	6.076347	6.868667	36.527458

Figure: Benchmarking of colSums2(), .colSums(), colSums() and apply+sum() on double+1000x100 data as well as rowSums2(), .rowSums(), rowSums() and apply+sum() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colSums2() and rowSums2() on double+1000x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	rowSums2	107.074	108.9395	127.9414	120.2215	137.0145	227.836
1	colSums2	105.267	111.0315	127.7037	126.3600	138.7345	186.246

	expr	min	lq	mean	median	uq	max
2	rowSums2	1.0000000	1.000000	1.0000000	1.00000	1.000000	1.0000000
1	colSums2	0.9831238	1.019203	0.9981425	1.05106	1.012553	0.8174564

Figure: Benchmarking of colSums2() and rowSums2() 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_0.60.0     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] rstudioapi_0.13         rappdirs_0.3.3          startup_0.15.0         
[67] labeling_0.4.2          bitops_1.0-7            base64enc_0.1-3        
[70] boot_1.3-28             gtable_0.3.0            DBI_1.1.1              
[73] markdown_1.1            R6_2.5.1                lpSolveAPI_5.5.2.0-17.7
[76] rle_0.9.2               dplyr_1.0.7             fastmap_1.1.0          
[79] bit_4.0.4               utf8_1.2.2              parallel_4.1.1         
[82] Rcpp_1.0.7              vctrs_0.3.8             png_0.1-7              
[85] DEoptimR_1.0-9          tidyselect_1.1.1        xfun_0.25              
[88] coda_0.19-4            

Total processing time was 26.44 secs.

Reproducibility

To reproduce this report, do:

html <- matrixStats:::benchmark('colSums2')

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 22:30:01 (+0200 UTC). Powered by RSP.