matrixStats.benchmarks

---

mean2() benchmarks on subsetted computation

This report benchmark the performance of mean2() on subsetted computation.

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)

Results

n = 1000 vector

> x <- data[["n = 1000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5343782 285.4    8529671 455.6   8529671 455.6
Vcells 16994035 129.7   33615701 256.5 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	mean2_x_S	0.002617	0.0026800	0.0027879	0.0027855	0.0028580	0.003139
2	mean2_x_S_no_refine	0.002625	0.0026925	0.0028074	0.0027875	0.0028820	0.003789
6	mean2_no_refine(x[idxs])	0.005270	0.0054265	0.0069546	0.0055245	0.0056670	0.145760
5	mean2(x[idxs])	0.005305	0.0054455	0.0055669	0.0055265	0.0056240	0.006907
4	mean2_no_refine(x, idxs)	0.005361	0.0054705	0.0056265	0.0055845	0.0057365	0.006292
3	mean2(x, idxs)	0.005354	0.0055135	0.0056912	0.0056400	0.0058250	0.007782

	expr	min	lq	mean	median	uq	max
1	mean2_x_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	mean2_x_S_no_refine	1.003057	1.004664	1.006962	1.000718	1.008398	1.207072
6	mean2_no_refine(x[idxs])	2.013756	2.024813	2.494516	1.983306	1.982855	46.435170
5	mean2(x[idxs])	2.027130	2.031903	1.996786	1.984024	1.967810	2.200382
4	mean2_no_refine(x, idxs)	2.048529	2.041231	2.018142	2.004847	2.007173	2.004460
3	mean2(x, idxs)	2.045854	2.057276	2.041364	2.024771	2.038139	2.479134

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 1000 data. Outliers are displayed as crosses. Times are in milliseconds.

n = 10000 vector

> x <- data[["n = 10000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5331508 284.8    8529671 455.6   8529671 455.6
Vcells 15833074 120.8   33615701 256.5 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 10000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	mean2_x_S	0.010794	0.0114820	0.0122459	0.0122445	0.0127205	0.013998
2	mean2_x_S_no_refine	0.010723	0.0114335	0.0123639	0.0123120	0.0130385	0.015634
6	mean2_no_refine(x[idxs])	0.029503	0.0311895	0.0333866	0.0328465	0.0348800	0.046098
5	mean2(x[idxs])	0.028908	0.0317200	0.0335386	0.0335125	0.0349180	0.049548
4	mean2_no_refine(x, idxs)	0.030674	0.0328510	0.0362676	0.0351895	0.0389210	0.070046
3	mean2(x, idxs)	0.030257	0.0322325	0.0359282	0.0357235	0.0380070	0.055797

	expr	min	lq	mean	median	uq	max
1	mean2_x_S	1.0000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	mean2_x_S_no_refine	0.9934223	0.995776	1.009643	1.005513	1.024999	1.116874
6	mean2_no_refine(x[idxs])	2.7332777	2.716382	2.726360	2.682551	2.742031	3.293185
5	mean2(x[idxs])	2.6781545	2.762585	2.738773	2.736943	2.745018	3.539649
4	mean2_no_refine(x, idxs)	2.8417639	2.861087	2.961620	2.873903	3.059707	5.004001
3	mean2(x, idxs)	2.8031314	2.807220	2.933910	2.917514	2.987854	3.986069

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 10000 data. Outliers are displayed as crosses. Times are in milliseconds.

n = 100000 vector

> x <- data[["n = 100000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5331607 284.8    8529671 455.6   8529671 455.6
Vcells 15896652 121.3   33615701 256.5 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 100000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	0.073527	0.0756660	0.0851851	0.0783055	0.0925870	0.129134
1	mean2_x_S	0.073509	0.0756385	0.0837340	0.0794795	0.0897670	0.117633
6	mean2_no_refine(x[idxs])	0.221539	0.2335015	0.2773580	0.2771705	0.3177650	0.374438
5	mean2(x[idxs])	0.221734	0.2386475	0.2804623	0.2785185	0.3143685	0.457560
3	mean2(x, idxs)	0.249248	0.2649270	0.3884439	0.2992800	0.4184140	6.716517
4	mean2_no_refine(x, idxs)	0.245279	0.2639460	0.3185506	0.3024775	0.3546150	0.434245

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	1.0000000	1.0000000	1.0000000	1.000000	1.0000000	1.0000000
1	mean2_x_S	0.9997552	0.9996366	0.9829651	1.014993	0.9695422	0.9109375
6	mean2_no_refine(x[idxs])	3.0130292	3.0859501	3.2559449	3.539604	3.4320693	2.8996082
5	mean2(x[idxs])	3.0156813	3.1539595	3.2923877	3.556819	3.3953849	3.5432961
3	mean2(x, idxs)	3.3898840	3.5012687	4.5599985	3.821954	4.5191442	52.0119953
4	mean2_no_refine(x, idxs)	3.3359038	3.4883039	3.7395113	3.862787	3.8300733	3.3627472

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 100000 data. Outliers are displayed as crosses. Times are in milliseconds.

n = 1000000 vector

> x <- data[["n = 1000000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5331706 284.8    8529671 455.6   8529671 455.6
Vcells 16526934 126.1   33615701 256.5 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 1000000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	mean2_x_S	0.888980	0.970562	0.9870926	0.9813695	1.004870	1.153251
2	mean2_x_S_no_refine	0.908760	0.970577	0.9978088	0.9825275	1.019047	1.260782
6	mean2_no_refine(x[idxs])	3.841791	4.892468	5.3651200	5.0987125	5.278500	16.702025
5	mean2(x[idxs])	3.957112	4.925172	5.1072061	5.1187365	5.339589	6.302222
3	mean2(x, idxs)	4.279667	6.379384	6.5634663	6.5185200	6.940049	16.112859
4	mean2_no_refine(x, idxs)	4.347253	6.408241	7.2835367	6.6077075	7.058704	17.249946

	expr	min	lq	mean	median	uq	max
1	mean2_x_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
2	mean2_x_S_no_refine	1.022250	1.000015	1.010856	1.001180	1.014108	1.093242
6	mean2_no_refine(x[idxs])	4.321572	5.040860	5.435275	5.195507	5.252919	14.482558
5	mean2(x[idxs])	4.451295	5.074556	5.173989	5.215911	5.313711	5.464744
3	mean2(x, idxs)	4.814132	6.572876	6.649292	6.642269	6.906415	13.971684
4	mean2_no_refine(x, idxs)	4.890158	6.602609	7.378778	6.733149	7.024495	14.957668

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 1000000 data. Outliers are displayed as crosses. Times are in milliseconds.

n = 10000000 vector

> x <- data[["n = 10000000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5331805 284.8    8529671 455.6   8529671 455.6
Vcells 22827611 174.2   40418841 308.4 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 10000000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	mean2_x_S	8.963665	12.25291	13.42115	13.53164	14.27329	17.88091
2	mean2_x_S_no_refine	9.692483	12.69796	13.90028	13.56586	16.41305	17.10707
3	mean2(x, idxs)	116.771360	126.50532	132.13719	133.08509	137.16898	144.61707
4	mean2_no_refine(x, idxs)	117.632574	128.45040	133.48622	133.49306	139.71935	147.40210
6	mean2_no_refine(x[idxs])	124.669728	135.14550	143.87622	141.00332	144.38891	523.71038
5	mean2(x[idxs])	124.945997	136.12857	140.59447	141.01150	144.73820	153.55588

	expr	min	lq	mean	median	uq	max
1	mean2_x_S	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
2	mean2_x_S_no_refine	1.081308	1.036323	1.035700	1.002529	1.149913	0.9567224
3	mean2(x, idxs)	13.027189	10.324515	9.845444	9.835104	9.610185	8.0877901
4	mean2_no_refine(x, idxs)	13.123268	10.483260	9.945958	9.865253	9.788866	8.2435443
6	mean2_no_refine(x[idxs])	13.908343	11.029669	10.720110	10.420268	10.116020	29.2887940
5	mean2(x[idxs])	13.939164	11.109901	10.475589	10.420872	10.140491	8.5876980

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on integer+n = 10000000 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)

Results

n = 1000 vector

> x <- data[["n = 1000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5331907 284.8    8529671 455.6   8529671 455.6
Vcells 21384601 163.2   40418841 308.4 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	0.002801	0.0028840	0.0032010	0.0029540	0.0030335	0.017288
1	mean2_x_S	0.003950	0.0040340	0.0041943	0.0040870	0.0042330	0.010007
6	mean2_no_refine(x[idxs])	0.005062	0.0053085	0.0057585	0.0054570	0.0056485	0.023367
4	mean2_no_refine(x, idxs)	0.005434	0.0056365	0.0058133	0.0057765	0.0059000	0.006759
5	mean2(x[idxs])	0.006200	0.0064090	0.0066228	0.0065635	0.0067735	0.007747
3	mean2(x, idxs)	0.006665	0.0068155	0.0070322	0.0069650	0.0071795	0.009142

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	1.000000	1.000000	1.000000	1.000000	1.000000	1.0000000
1	mean2_x_S	1.410211	1.398752	1.310285	1.383548	1.395418	0.5788408
6	mean2_no_refine(x[idxs])	1.807212	1.840673	1.798955	1.847326	1.862040	1.3516312
4	mean2_no_refine(x, idxs)	1.940021	1.954404	1.816084	1.955484	1.944948	0.3909648
5	mean2(x[idxs])	2.213495	2.222261	2.068971	2.221903	2.232899	0.4481143
3	mean2(x, idxs)	2.379507	2.363211	2.196858	2.357820	2.366738	0.5288061

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 1000 data. Outliers are displayed as crosses. Times are in milliseconds.

n = 10000 vector

> x <- data[["n = 10000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5332003 284.8    8529671 455.6   8529671 455.6
Vcells 21394560 163.3   40418841 308.4 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 10000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	0.013189	0.0136025	0.0146243	0.0143285	0.0153315	0.019333
1	mean2_x_S	0.022512	0.0229240	0.0248455	0.0244275	0.0261100	0.043589
6	mean2_no_refine(x[idxs])	0.029430	0.0310335	0.0334698	0.0323710	0.0358140	0.046852
4	mean2_no_refine(x, idxs)	0.032794	0.0334905	0.0372509	0.0355420	0.0391355	0.076246
5	mean2(x[idxs])	0.038999	0.0409790	0.0440330	0.0437330	0.0466370	0.059871
3	mean2(x, idxs)	0.043297	0.0444420	0.0484117	0.0470765	0.0514175	0.067115

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	mean2_x_S	1.706877	1.685278	1.698924	1.704819	1.703030	2.254642
6	mean2_no_refine(x[idxs])	2.231405	2.281456	2.288645	2.259204	2.335975	2.423421
4	mean2_no_refine(x, idxs)	2.486466	2.462084	2.547199	2.480511	2.552620	3.943827
5	mean2(x[idxs])	2.956934	3.012608	3.010952	3.052169	3.041907	3.096829
3	mean2(x, idxs)	3.282811	3.267193	3.310364	3.285515	3.353716	3.471525

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 10000 data. Outliers are displayed as crosses. Times are in milliseconds.

n = 100000 vector

> x <- data[["n = 100000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5332102 284.8    8529671 455.6   8529671 455.6
Vcells 21489668 164.0   40418841 308.4 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 100000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	0.091491	0.0919595	0.1028826	0.0960520	0.1081820	0.164275
1	mean2_x_S	0.163594	0.1655625	0.1798339	0.1725650	0.1915510	0.231438
4	mean2_no_refine(x, idxs)	0.269157	0.2968530	0.3619284	0.3521040	0.4270280	0.478045
6	mean2_no_refine(x[idxs])	0.284739	0.3253185	0.3742937	0.3837795	0.4244205	0.451456
5	mean2(x[idxs])	0.354564	0.4028905	0.4526132	0.4672430	0.4956540	0.749004
3	mean2(x, idxs)	0.393542	0.4639200	0.6129403	0.5469265	0.5587500	10.399488

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	mean2_x_S	1.788088	1.800385	1.747953	1.796579	1.770636	1.408845
4	mean2_no_refine(x, idxs)	2.941896	3.228084	3.517878	3.665764	3.947311	2.910029
6	mean2_no_refine(x[idxs])	3.112208	3.537628	3.638066	3.995539	3.923208	2.748172
5	mean2(x[idxs])	3.875398	4.381173	4.399318	4.864480	4.581668	4.559452
3	mean2(x, idxs)	4.301429	5.044829	5.957668	5.694067	5.164907	63.305360

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 100000 data. Outliers are displayed as crosses. Times are in milliseconds.

n = 1000000 vector

> x <- data[["n = 1000000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5332201 284.8    8529671 455.6   8529671 455.6
Vcells 22434735 171.2   40418841 308.4 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 1000000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	1.343081	1.549252	1.578564	1.576547	1.632972	1.918255
1	mean2_x_S	2.515176	2.804453	2.870944	2.876253	2.955961	3.323344
6	mean2_no_refine(x[idxs])	9.380240	10.324032	11.127897	10.477827	10.614484	21.379512
4	mean2_no_refine(x, idxs)	8.200189	10.458063	10.982563	10.562641	10.846580	29.819252
5	mean2(x[idxs])	11.320825	11.651995	12.412661	11.775655	11.939690	22.736827
3	mean2(x, idxs)	15.378069	15.684134	16.370287	15.831978	16.128079	27.183240

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
1	mean2_x_S	1.872691	1.810198	1.818706	1.824401	1.810172	1.732483
6	mean2_no_refine(x[idxs])	6.984121	6.663882	7.049381	6.646061	6.500099	11.145292
4	mean2_no_refine(x, idxs)	6.105506	6.750395	6.957314	6.699858	6.642231	15.544989
5	mean2(x[idxs])	8.428997	7.521046	7.863263	7.469270	7.311630	11.852870
3	mean2(x, idxs)	11.449845	10.123682	10.370368	10.042186	9.876516	14.170817

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 1000000 data. Outliers are displayed as crosses. Times are in milliseconds.

n = 10000000 vector

> x <- data[["n = 10000000"]]
> idxs <- sample.int(length(x), size = length(x) * 0.7)
> x_S <- x[idxs]
> gc()
           used  (Mb) gc trigger  (Mb)  max used  (Mb)
Ncells  5332294 284.8    8529671 455.6   8529671 455.6
Vcells 31885445 243.3   48582609 370.7 101881463 777.3
> stats <- microbenchmark(mean2_x_S = mean2(x_S, refine = TRUE), mean2_x_S_no_refine = mean2(x_S, refine = FALSE), 
+     `mean2(x, idxs)` = mean2(x, idxs = idxs, refine = TRUE), `mean2_no_refine(x, idxs)` = mean2(x, 
+         idxs = idxs, refine = FALSE), `mean2(x[idxs])` = mean2(x[idxs], refine = TRUE), `mean2_no_refine(x[idxs])` = mean2(x[idxs], 
+         refine = FALSE), unit = "ms")

Table: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 10000000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	10.88876	14.49473	17.07083	15.97592	21.50374	24.61304
1	mean2_x_S	18.75323	23.55524	27.70298	25.66344	31.49576	39.18378
4	mean2_no_refine(x, idxs)	148.64750	170.24358	184.80863	183.40903	190.31554	558.94832
6	mean2_no_refine(x[idxs])	169.73511	179.18399	188.45000	184.48299	196.81973	215.57362
5	mean2(x[idxs])	178.75786	195.58341	208.37115	206.84077	210.96170	572.83587
3	mean2(x, idxs)	281.16566	311.13864	321.84314	322.05485	334.64647	357.05475

	expr	min	lq	mean	median	uq	max
2	mean2_x_S_no_refine	1.000000	1.00000	1.000000	1.000000	1.000000	1.000000
1	mean2_x_S	1.722256	1.62509	1.622826	1.606383	1.464665	1.591993
4	mean2_no_refine(x, idxs)	13.651464	11.74521	10.825991	11.480344	8.850348	22.709440
6	mean2_no_refine(x[idxs])	15.588104	12.36201	11.039300	11.547568	9.152815	8.758513
5	mean2(x[idxs])	16.416734	13.49342	12.206270	12.947036	9.810467	23.273675
3	mean2(x, idxs)	25.821644	21.46564	18.853399	20.158771	15.562248	14.506731

Figure: Benchmarking of mean2_x_S(), mean2_x_S_no_refine(), mean2(x, idxs)(), mean2_no_refine(x, idxs)(), mean2(x[idxs])() and mean2_no_refine(x[idxs])() on double+n = 10000000 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 2.9 mins.

Reproducibility

To reproduce this report, do:

html <- matrixStats:::benchmark('mean2_subset')

---

Copyright Dongcan Jiang. Last updated on 2021-08-25 19:23:16 (+0200 UTC). Powered by RSP.