matrixStats.benchmarks

---

allocVector() benchmarks

This report benchmark the performance of allocVector() against alternative methods.

Alternative methods

• vector() + assignment
• rep()
• matrix() + as.vector()

where

> allocVector_R1 <- function(length, value = NA) {
+     x <- vector(mode = typeof(value), length = length)
+     if (!is.finite(value) || value != 0) 
+         x[] <- value
+     x
+ }
> allocVector_R2 <- function(length, value = NA) {
+     x <- matrix(data = value, nrow = length, ncol = 1L)
+     as.vector(x)
+ }

Data type “integer”

Data

> rvector <- function(n, mode = c("logical", "double", "integer"), range = c(-100, +100), na_prob = 0) {
+     mode <- match.arg(mode)
+     if (mode == "logical") {
+         x <- sample(c(FALSE, TRUE), size = n, replace = TRUE)
+     }     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
+     x
+ }
> rvectors <- function(scale = 10, seed = 1, ...) {
+     set.seed(seed)
+     data <- list()
+     data[[1]] <- rvector(n = scale * 100, ...)
+     data[[2]] <- rvector(n = scale * 1000, ...)
+     data[[3]] <- rvector(n = scale * 10000, ...)
+     data[[4]] <- rvector(n = scale * 1e+05, ...)
+     data[[5]] <- rvector(n = scale * 1e+06, ...)
+     names(data) <- sprintf("n = %d", sapply(data, FUN = length))
+     data
+ }
> data <- rvectors(mode = mode)
> values <- list(zero = 0, one = 1, `NA` = NA_real_)
> if (mode != "double") 
+     values <- lapply(values, FUN = function(x) {
+         storage.mode(x) <- mode
+         x
+     })

Results

n = 1000 matrix

> n <- length(data[["n = 1000"]])
> str(value)
 int 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.001291	0.0016485	0.0025727	0.0017720	0.0019845	0.077012
2	rep	0.002220	0.0023415	0.0025900	0.0024800	0.0025825	0.009392
3	allocVector_R1	0.002845	0.0032365	0.0036185	0.0034215	0.0036040	0.018481
4	allocVector_R2	0.004760	0.0051805	0.0057465	0.0054070	0.0059185	0.020808

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rep	1.719597	1.420382	1.006705	1.399549	1.301335	0.1219550
3	allocVector_R1	2.203718	1.963300	1.406462	1.930869	1.816075	0.2399756
4	allocVector_R2	3.687064	3.142554	2.233603	3.051354	2.982363	0.2701917

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 1000"]])
> str(value)
 int 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.001608	0.0019880	0.0022513	0.0021310	0.0023155	0.009959
2	rep	0.002176	0.0023925	0.0025901	0.0025400	0.0026475	0.004491
4	allocVector_R2	0.004733	0.0053460	0.0059788	0.0056220	0.0059840	0.029449
3	allocVector_R1	0.006413	0.0069215	0.0074955	0.0071845	0.0076400	0.021647

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rep	1.353234	1.203471	1.150483	1.191929	1.143382	0.4509489
4	allocVector_R2	2.943408	2.689135	2.655747	2.638198	2.584323	2.9570238
3	allocVector_R1	3.988184	3.481640	3.329435	3.371422	3.299503	2.1736118

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 1000"]])
> str(value)
 int NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.001671	0.0019900	0.0023489	0.0021735	0.0023745	0.013166
2	rep	0.002120	0.0024015	0.0027203	0.0024985	0.0026735	0.019697
4	allocVector_R2	0.004969	0.0054450	0.0059545	0.0057455	0.0060480	0.021023
3	allocVector_R1	0.006471	0.0068795	0.0075014	0.0070935	0.0073740	0.033466

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rep	1.268701	1.206784	1.158105	1.149528	1.125921	1.496050
4	allocVector_R2	2.973669	2.736181	2.534990	2.643432	2.547063	1.596764
3	allocVector_R1	3.872531	3.457035	3.193517	3.263630	3.105496	2.541850

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

n = 10000 matrix

> n <- length(data[["n = 10000"]])
> str(value)
 int 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.003047	0.0034870	0.0038539	0.0037400	0.0039890	0.011583
3	allocVector_R1	0.004705	0.0051640	0.0055503	0.0054305	0.0057080	0.013387
4	allocVector_R2	0.011985	0.0132340	0.0143255	0.0137170	0.0143160	0.045671
2	rep	0.016343	0.0164935	0.0167819	0.0166105	0.0167905	0.024313

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	allocVector_R1	1.544142	1.480929	1.440160	1.452005	1.430935	1.155746
4	allocVector_R2	3.933377	3.795240	3.717135	3.667647	3.588869	3.942934
2	rep	5.363636	4.729997	4.354512	4.441310	4.209200	2.099024

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 10000"]])
> str(value)
 int 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.006401	0.0072415	0.0075306	0.0074175	0.0076965	0.015085
4	allocVector_R2	0.012174	0.0135805	0.0146636	0.0141695	0.0147865	0.030900
2	rep	0.016428	0.0165950	0.0169292	0.0167140	0.0169350	0.021704
3	allocVector_R1	0.031605	0.0322645	0.0331387	0.0325395	0.0328135	0.064134

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	allocVector_R2	1.901890	1.875371	1.947207	1.910280	1.921198	2.048392
2	rep	2.566474	2.291652	2.248073	2.253320	2.200351	1.438780
3	allocVector_R1	4.937510	4.455500	4.400568	4.386855	4.263431	4.251508

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 10000"]])
> str(value)
 int NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.006388	0.0069675	0.0074947	0.0073320	0.0077120	0.015878
4	allocVector_R2	0.012178	0.0133495	0.0143775	0.0140195	0.0147885	0.030996
2	rep	0.016430	0.0165820	0.0174405	0.0167040	0.0170115	0.058788
3	allocVector_R1	0.031339	0.0321005	0.0329580	0.0325600	0.0332745	0.049945

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	allocVector_R2	1.906387	1.915967	1.918352	1.912098	1.917596	1.952135
2	rep	2.572010	2.379907	2.327028	2.278232	2.205848	3.702481
3	allocVector_R1	4.905917	4.607176	4.397489	4.440807	4.314640	3.145547

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

n = 100000 matrix

> n <- length(data[["n = 100000"]])
> str(value)
 int 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 100000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.017483	0.0202825	0.0758000	0.0237475	0.1618945	0.183558
3	allocVector_R1	0.018691	0.0215060	0.0726146	0.0247675	0.1617020	0.211468
4	allocVector_R2	0.076114	0.0860500	0.2589109	0.1028640	0.3474015	5.896655
2	rep	0.126403	0.1339245	0.2467912	0.1584090	0.2577160	5.898492

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.0000000	1.000000	1.000000	1.00000
3	allocVector_R1	1.069096	1.060323	0.9579765	1.042952	0.998811	1.15205
4	allocVector_R2	4.353601	4.242574	3.4157123	4.331572	2.145851	32.12421
2	rep	7.230052	6.602958	3.2558221	6.670555	1.591876	32.13421

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 100000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 100000"]])
> str(value)
 int 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 100000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.043691	0.0456425	0.1471062	0.0535520	0.1681350	5.466483
4	allocVector_R2	0.075244	0.0849685	0.2287936	0.0977260	0.3340995	5.711340
2	rep	0.123369	0.1276035	0.1793324	0.1524165	0.2520245	0.293475
3	allocVector_R1	0.216631	0.2270355	0.3464223	0.2790740	0.4904475	0.571319

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
4	allocVector_R2	1.722185	1.861609	1.555295	1.824880	1.987091	1.0447924
2	rep	2.823671	2.795717	1.219067	2.846140	1.498941	0.0536863
3	allocVector_R1	4.958252	4.974213	2.354913	5.211271	2.916986	0.1045131

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 100000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 100000"]])
> str(value)
 int NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 100000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.043393	0.0457640	0.0841466	0.0501390	0.1605895	0.201411
4	allocVector_R2	0.074571	0.0861125	0.1791867	0.1002435	0.3397675	0.395409
2	rep	0.126574	0.1321670	0.2319547	0.1524180	0.2403570	5.846479
3	allocVector_R1	0.222513	0.2309980	0.3958427	0.2737885	0.4893180	6.102523

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	allocVector_R2	1.718503	1.881665	2.129459	1.999312	2.115752	1.963195
2	rep	2.916922	2.888012	2.756556	3.039909	1.496717	29.027605
3	allocVector_R1	5.127855	5.047592	4.704206	5.460590	3.047011	30.298857

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 100000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

n = 1000000 matrix

> n <- length(data[["n = 1000000"]])
> str(value)
 int 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.167854	0.1932035	0.6314815	0.2049925	0.2340575	6.194362
3	allocVector_R1	0.175405	0.2003095	0.6118296	0.2158965	0.2373765	4.106948
4	allocVector_R2	0.919935	0.9767010	1.5225728	1.0038025	1.0480050	6.820040
2	rep	1.218644	1.3482700	1.6845087	1.4079745	1.4658250	7.524439

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
3	allocVector_R1	1.044985	1.036780	0.9688796	1.053192	1.014180	0.6630139
4	allocVector_R2	5.480566	5.055297	2.4111122	4.896777	4.477554	1.1010077
2	rep	7.260143	6.978497	2.6675501	6.868420	6.262670	1.2147238

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 1000000"]])
> str(value)
 int 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.412375	0.4283630	0.891144	0.4389290	0.4659515	5.902721
4	allocVector_R2	0.872616	0.9157895	1.467150	0.9277395	0.9500330	6.577447
2	rep	1.167224	1.2292450	1.416293	1.2601560	1.2729535	4.918317
3	allocVector_R1	2.086407	2.2512220	2.935449	2.2847845	2.3412995	7.085598

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
4	allocVector_R2	2.116074	2.137882	1.646367	2.113644	2.038910	1.1143076
2	rep	2.830492	2.869634	1.589298	2.870979	2.731944	0.8332288
3	allocVector_R1	5.059489	5.255407	3.294023	5.205362	5.024771	1.2003952

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 1000000"]])
> str(value)
 int NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.412432	0.4321405	0.8478479	0.437984	0.5387145	6.619624
4	allocVector_R2	0.889459	0.9166540	1.5997440	0.930059	1.0314850	6.800204
2	rep	1.183535	1.2285340	5.0905064	1.260788	1.3288735	341.470976
3	allocVector_R1	2.186592	2.2412955	3.0821365	2.285582	3.6176760	8.634266

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
4	allocVector_R2	2.156620	2.121194	1.886829	2.123500	1.914716	1.027279
2	rep	2.869649	2.842904	6.004033	2.878617	2.466749	51.584648
3	allocVector_R1	5.301703	5.186497	3.635247	5.218414	6.715386	1.304344

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 1000000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

n = 10000000 matrix

> n <- length(data[["n = 10000000"]])
> str(value)
 int 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
3	allocVector_R1	12.63546	13.02242	20.16692	13.65514	21.85036	351.92373
1	allocVector	12.65063	13.02751	16.89038	13.76262	21.80091	23.74778
2	rep	20.45499	20.86397	32.21169	22.45490	30.39022	366.81992
4	allocVector_R2	29.58066	36.67542	40.33890	38.04415	39.47954	377.51608

	expr	min	lq	mean	median	uq	max
3	allocVector_R1	1.000000	1.000000	1.000000	1.000000	1.0000000	1.0000000
1	allocVector	1.001201	1.000391	0.837529	1.007871	0.9977372	0.0674799
2	rep	1.618856	1.602157	1.597254	1.644428	1.3908342	1.0423279
4	allocVector_R2	2.341082	2.816329	2.000251	2.786068	1.8068146	1.0727213

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 10000000"]])
> str(value)
 int 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	13.23932	13.64579	20.96921	14.32506	22.39856	357.4252
2	rep	20.34816	21.12817	28.70912	22.76576	29.92917	357.9072
4	allocVector_R2	29.65705	36.88882	51.00195	38.35922	39.49794	372.1699
3	allocVector_R1	40.42712	47.83018	55.29098	48.88169	50.40713	394.9465

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rep	1.536948	1.548329	1.369109	1.589226	1.336210	1.001348
4	allocVector_R2	2.240073	2.703312	2.432230	2.677771	1.763414	1.041252
3	allocVector_R1	3.053564	3.505124	2.636770	3.412321	2.250463	1.104977

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 10000000"]])
> str(value)
 int NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	13.20015	13.77348	20.71589	14.19909	22.31642	359.7546
2	rep	20.41913	21.61964	35.40740	22.77208	30.07727	357.6511
4	allocVector_R2	29.49000	36.33405	43.83292	38.06939	39.66189	368.1529
3	allocVector_R1	40.62559	48.08906	58.71423	49.56928	50.75201	387.2327

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rep	1.546887	1.569657	1.709190	1.603770	1.347764	0.9941532
4	allocVector_R2	2.234066	2.637972	2.115908	2.681114	1.777252	1.0233448
3	allocVector_R1	3.077661	3.491425	2.834260	3.491017	2.274201	1.0763802

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on integer+n = 10000000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

Data type “double”

Data

> rvector <- function(n, mode = c("logical", "double", "integer"), range = c(-100, +100), na_prob = 0) {
+     mode <- match.arg(mode)
+     if (mode == "logical") {
+         x <- sample(c(FALSE, TRUE), size = n, replace = TRUE)
+     }     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
+     x
+ }
> rvectors <- function(scale = 10, seed = 1, ...) {
+     set.seed(seed)
+     data <- list()
+     data[[1]] <- rvector(n = scale * 100, ...)
+     data[[2]] <- rvector(n = scale * 1000, ...)
+     data[[3]] <- rvector(n = scale * 10000, ...)
+     data[[4]] <- rvector(n = scale * 1e+05, ...)
+     data[[5]] <- rvector(n = scale * 1e+06, ...)
+     names(data) <- sprintf("n = %d", sapply(data, FUN = length))
+     data
+ }
> data <- rvectors(mode = mode)
> values <- list(zero = 0, one = 1, `NA` = NA_real_)
> if (mode != "double") 
+     values <- lapply(values, FUN = function(x) {
+         storage.mode(x) <- mode
+         x
+     })

Results

n = 1000 matrix

> n <- length(data[["n = 1000"]])
> str(value)
 num 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.001420	0.0015895	0.0019484	0.0017310	0.0018795	0.015117
2	rep	0.002024	0.0021010	0.0022897	0.0021690	0.0023630	0.004466
3	allocVector_R1	0.002876	0.0030970	0.0034427	0.0033025	0.0035025	0.012204
4	allocVector_R2	0.004793	0.0050725	0.0055243	0.0052555	0.0055090	0.023478

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rep	1.425352	1.321799	1.175172	1.253033	1.257249	0.295429
3	allocVector_R1	2.025352	1.948411	1.766884	1.907857	1.863528	0.807303
4	allocVector_R2	3.375352	3.191255	2.835257	3.036106	2.931099	1.553086

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 1000"]])
> str(value)
 num 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.001681	0.002004	0.0023142	0.0021520	0.0022920	0.014470
2	rep	0.002105	0.002248	0.0024207	0.0023670	0.0024815	0.004899
4	allocVector_R2	0.004855	0.005267	0.0059615	0.0054860	0.0057430	0.038419
3	allocVector_R1	0.006310	0.006711	0.0072224	0.0069395	0.0071400	0.028761

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rep	1.252231	1.121757	1.046016	1.099907	1.082679	0.3385625
4	allocVector_R2	2.888162	2.628243	2.576061	2.549256	2.505672	2.6550795
3	allocVector_R1	3.753718	3.348802	3.120924	3.224675	3.115183	1.9876296

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 1000"]])
> str(value)
 num NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.001828	0.0021695	0.0024641	0.002324	0.0025125	0.013616
2	rep	0.002124	0.0022440	0.0024270	0.002336	0.0025085	0.006572
4	allocVector_R2	0.005172	0.0054750	0.0058928	0.005744	0.0059930	0.014087
3	allocVector_R1	0.006307	0.0066490	0.0072830	0.006839	0.0071050	0.037852

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	rep	1.161926	1.034340	0.9849598	1.005164	0.998408	0.4826675
4	allocVector_R2	2.829322	2.523623	2.3915083	2.471601	2.385274	1.0345917
3	allocVector_R1	3.450219	3.064761	2.9556748	2.942771	2.827861	2.7799647

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

n = 10000 matrix

> n <- length(data[["n = 10000"]])
> str(value)
 num 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.005112	0.0056390	0.0073907	0.0060030	0.006675	0.030509
3	allocVector_R1	0.006537	0.0072380	0.0090508	0.0077700	0.008711	0.029767
2	rep	0.016425	0.0167215	0.0188860	0.0168590	0.017179	0.041688
4	allocVector_R2	0.016705	0.0181205	0.0247799	0.0214445	0.023861	0.071965

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	allocVector_R1	1.278756	1.283561	1.224625	1.294353	1.305019	0.9756793
2	rep	3.213028	2.965331	2.555383	2.808429	2.573633	1.3664165
4	allocVector_R2	3.267801	3.213424	3.352864	3.572297	3.574682	2.3588122

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 10000"]])
> str(value)
 num 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.008087	0.0087930	0.0096728	0.0091775	0.0099930	0.020400
2	rep	0.015890	0.0161675	0.0169276	0.0167310	0.0170655	0.030499
4	allocVector_R2	0.016701	0.0179160	0.0194537	0.0188615	0.0205270	0.030708
3	allocVector_R1	0.029915	0.0309650	0.0323190	0.0319640	0.0324835	0.064344

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rep	1.964882	1.838679	1.750012	1.823046	1.707745	1.495049
4	allocVector_R2	2.065166	2.037530	2.011165	2.055189	2.054138	1.505294
3	allocVector_R1	3.699147	3.521551	3.341218	3.482866	3.250625	3.154118

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 10000"]])
> str(value)
 num NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.008084	0.0087125	0.0097897	0.0091590	0.0110350	0.016765
2	rep	0.015933	0.0165005	0.0170365	0.0168925	0.0170990	0.029958
4	allocVector_R2	0.016285	0.0180090	0.0199812	0.0191240	0.0214535	0.033657
3	allocVector_R1	0.028315	0.0310975	0.0321950	0.0317230	0.0322925	0.064343

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rep	1.970930	1.893888	1.740258	1.844361	1.549524	1.786937
4	allocVector_R2	2.014473	2.067030	2.041047	2.088001	1.944132	2.007575
3	allocVector_R1	3.502598	3.569297	3.288673	3.463588	2.926371	3.837936

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

n = 100000 matrix

> n <- length(data[["n = 100000"]])
> str(value)
 num 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 100000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
3	allocVector_R1	0.032641	0.0449265	0.2199961	0.2854820	0.3175340	0.451317
1	allocVector	0.033243	0.0456940	0.2229439	0.2867635	0.3176900	0.411610
2	rep	0.123580	0.1597170	0.4444608	0.3618950	0.3876625	7.486867
4	allocVector_R2	0.124075	0.1720470	0.4737744	0.5797065	0.6271285	0.759641

	expr	min	lq	mean	median	uq	max
3	allocVector_R1	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	allocVector	1.018443	1.017084	1.013399	1.004489	1.000491	0.9120197
2	rep	3.786036	3.555073	2.020312	1.267663	1.220854	16.5889319
4	allocVector_R2	3.801201	3.829522	2.153558	2.030624	1.974996	1.6831650

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 100000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 100000"]])
> str(value)
 num 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 100000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.058183	0.0675595	0.1927077	0.2527905	0.2886095	0.357557
2	rep	0.120672	0.1300380	0.2791915	0.3255000	0.3699000	0.492998
3	allocVector_R1	0.210113	0.2214615	0.4975119	0.5187295	0.5849955	7.302006
4	allocVector_R2	0.125831	0.1801910	0.5903175	0.5495205	0.6041575	7.363703

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rep	2.074008	1.924792	1.448782	1.287627	1.281663	1.378796
3	allocVector_R1	3.611244	3.278022	2.581692	2.052013	2.026945	20.421936
4	allocVector_R2	2.162676	2.667145	3.063280	2.173818	2.093339	20.594487

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 100000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 100000"]])
> str(value)
 num NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 100000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.056891	0.0681750	0.3266920	0.2546510	0.2715175	7.396849
2	rep	0.119781	0.1594865	0.2848875	0.3340025	0.3772695	0.491985
3	allocVector_R1	0.203552	0.2210220	0.4985277	0.5216630	0.5798600	7.137726
4	allocVector_R2	0.124222	0.1666655	0.4480961	0.5427785	0.6066960	0.748163

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.0000000	1.000000	1.000000	1.0000000
2	rep	2.105447	2.339369	0.8720372	1.311609	1.389485	0.0665128
3	allocVector_R1	3.577930	3.241980	1.5259872	2.048541	2.135627	0.9649685
4	allocVector_R2	2.183509	2.444672	1.3716166	2.131460	2.234464	0.1011462

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 100000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

n = 1000000 matrix

> n <- length(data[["n = 1000000"]])
> str(value)
 num 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.295625	0.3784000	1.051034	0.4125735	0.4669035	7.638469
3	allocVector_R1	0.308285	0.3912355	1.113137	0.4248760	0.4791795	7.625590
2	rep	1.126853	1.3289500	1.967328	1.4070430	1.5321590	10.030630
4	allocVector_R2	1.334786	1.4031215	2.820345	1.4552420	5.1149475	9.890856

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
3	allocVector_R1	1.042825	1.033921	1.059088	1.029819	1.026292	0.9983139
2	rep	3.811765	3.512024	1.871802	3.410406	3.281532	1.3131728
4	allocVector_R2	4.515132	3.708038	2.683401	3.527231	10.955042	1.2948741

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 1000000"]])
> str(value)
 num 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.632168	0.6591875	1.486444	0.6705945	2.472943	7.413477
2	rep	1.095766	1.2298460	1.836087	1.2723820	1.362015	7.886881
4	allocVector_R2	1.327035	1.3697150	3.008941	1.4180105	5.045395	8.010199
3	allocVector_R1	1.986027	2.2137430	3.007227	2.2749715	2.473757	9.414505

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.0000000	1.000000
2	rep	1.733346	1.865700	1.235222	1.897394	0.5507667	1.063857
4	allocVector_R2	2.099181	2.077884	2.024255	2.114557	2.0402389	1.080491
3	allocVector_R1	3.141613	3.358290	2.023102	3.392470	1.0003290	1.269918

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 1000000"]])
> str(value)
 num NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	0.642240	0.6611895	1.286497	0.671658	0.7239615	7.644119
2	rep	1.113877	1.2673480	1.963661	1.278311	1.4059975	9.011668
4	allocVector_R2	1.322725	1.3697685	2.327147	1.408396	1.4762055	8.902168
3	allocVector_R1	2.007035	2.2587050	6.273910	2.293862	2.4749530	333.149908

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rep	1.734363	1.916770	1.526362	1.903217	1.942089	1.178902
4	allocVector_R2	2.059549	2.071673	1.808901	2.096895	2.039066	1.164577
3	allocVector_R1	3.125055	3.416123	4.876738	3.415224	3.418625	43.582512

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 1000000+NA data. Outliers are displayed as crosses. Times are in milliseconds.

n = 10000000 matrix

> n <- length(data[["n = 10000000"]])
> str(value)
 num 0

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000000+0 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	25.23340	25.79592	30.07851	25.93894	37.23915	41.25383
3	allocVector_R1	25.24707	25.81174	34.12057	26.01175	38.40659	367.32780
2	rep	31.19824	32.17893	44.82133	34.71728	44.19895	382.63226
4	allocVector_R2	56.31461	57.69006	68.05523	66.64490	69.04198	396.01130

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
3	allocVector_R1	1.000542	1.000613	1.134383	1.002807	1.031350	8.904090
2	rep	1.236387	1.247443	1.490144	1.338423	1.186895	9.275073
4	allocVector_R2	2.231749	2.236403	2.262586	2.569299	1.854016	9.599383

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000000+0 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 10000000"]])
> str(value)
 num 1

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000000+1 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	23.76478	25.04440	36.15933	25.73830	35.88860	376.01751
2	rep	31.52511	33.11105	44.20774	34.20348	42.48701	385.53139
3	allocVector_R1	52.29072	54.85647	64.46291	63.58004	66.20788	391.70816
4	allocVector_R2	56.47837	63.21143	65.77525	66.77103	69.58766	74.25866

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	rep	1.326548	1.322094	1.222582	1.328894	1.183858	1.0253017
3	allocVector_R1	2.200346	2.190369	1.782746	2.470250	1.844816	1.0417285
4	allocVector_R2	2.376558	2.523974	1.819040	2.594228	1.938991	0.1974872

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000000+1 data. Outliers are displayed as crosses. Times are in milliseconds.

> n <- length(data[["n = 10000000"]])
> str(value)
 num NA

> stats <- microbenchmark(allocVector = allocVector(length = n, value = value), rep = rep(value, times = n), 
+     allocVector_R1 = allocVector_R1(length = n, value = value), allocVector_R2 = allocVector_R2(length = n, 
+         value = value), unit = "ms")

Table: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000000+NA data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	allocVector	23.99955	25.62834	29.40072	26.25675	35.99857	41.97127
2	rep	31.14532	34.46510	42.75304	35.15820	46.00730	381.78028
3	allocVector_R1	52.27612	54.96357	66.91250	56.73542	66.76330	395.68549
4	allocVector_R2	56.58167	59.60112	66.55866	68.39408	71.06541	79.10422

	expr	min	lq	mean	median	uq	max
1	allocVector	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	rep	1.297746	1.344804	1.454149	1.339016	1.278031	9.096228
3	allocVector_R1	2.178213	2.144641	2.275880	2.160794	1.854610	9.427531
4	allocVector_R2	2.357614	2.325594	2.263845	2.604819	1.974117	1.884723

Figure: Benchmarking of allocVector(), rep(), allocVector_R1() and allocVector_R2() on double+n = 10000000+NA data. 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 2.29 mins.

Reproducibility

To reproduce this report, do:

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

---

Copyright Henrik Bengtsson. Last updated on 2021-08-25 18:47:00 (+0200 UTC). Powered by RSP.