matrixStats.benchmarks
colMads() and rowMads() benchmarks
This report benchmark the performance of colMads() and rowMads() against alternative methods.
Alternative methods
- apply() + mad()
- colMads2() and rowMads2()
where rowMads2() and colMads2() are:
> rowMads2 <- function(x, const = 1.4826, na.rm = FALSE) {
+ mu <- rowMedians(x, na.rm = na.rm)
+ x <- abs(x - mu)
+ mad <- rowMedians(x, na.rm = FALSE)
+ const * mad
+ }
> colMads2 <- function(x, const = 1.4826, na.rm = FALSE) {
+ mu <- colMedians(x, na.rm = na.rm)
+ x <- abs(x - mu)
+ mad <- colMedians(x, na.rm = FALSE)
+ const * mad
+ }
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 5236834 279.7 8529671 455.6 8529671 455.6
Vcells 10094224 77.1 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5227378 279.2 8529671 455.6 8529671 455.6
Vcells 10063140 76.8 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() 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 | colMads | 0.002911 | 0.0043845 | 0.0064869 | 0.005844 | 0.0076305 | 0.023153 |
| 2 | colMads2 | 0.005850 | 0.0081145 | 0.0119110 | 0.010377 | 0.0133185 | 0.083362 |
| 3 | apply+mad | 0.523131 | 0.5753125 | 0.6591002 | 0.647018 | 0.7109160 | 1.168096 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | colMads2 | 2.009619 | 1.850724 | 1.836173 | 1.775667 | 1.745429 | 3.600484 |
| 3 | apply+mad | 179.708348 | 131.215076 | 101.605422 | 110.714921 | 93.167682 | 50.451173 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 0.003372 | 0.0049500 | 0.0066732 | 0.0063460 | 0.007798 | 0.022042 |
| 2 | rowMads2 | 0.005950 | 0.0079055 | 0.0115473 | 0.0103530 | 0.012708 | 0.093851 |
| 3 | apply+mad | 0.518819 | 0.5595975 | 0.6468266 | 0.6272185 | 0.692554 | 1.020574 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.00000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.764531 | 1.597071 | 1.73041 | 1.631421 | 1.629649 | 4.257826 |
| 3 | apply+mad | 153.860913 | 113.050000 | 96.92930 | 98.836826 | 88.811747 | 46.301334 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on integer+10x10 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() on integer+10x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 2.911 | 4.3845 | 6.48686 | 5.844 | 7.6305 | 23.153 |
| 2 | rowMads | 3.372 | 4.9500 | 6.67318 | 6.346 | 7.7980 | 22.042 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.0000 | 1.000000 | 1.0000000 |
| 2 | rowMads | 1.158365 | 1.128977 | 1.028723 | 1.0859 | 1.021951 | 0.9520149 |
Figure: Benchmarking of colMads() and rowMads() 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 5225974 279.1 8529671 455.6 8529671 455.6
Vcells 9679850 73.9 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5225968 279.1 8529671 455.6 8529671 455.6
Vcells 9684893 73.9 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on integer+100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 0.280159 | 0.3086925 | 0.3192907 | 0.3133360 | 0.3226325 | 0.482616 |
| 2 | colMads2 | 0.337539 | 0.3668050 | 0.3827254 | 0.3758095 | 0.3862170 | 0.625727 |
| 3 | apply+mad | 5.557610 | 6.0530400 | 6.4049605 | 6.2980390 | 6.4710520 | 15.261305 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.00000 | 1.000000 |
| 2 | colMads2 | 1.204812 | 1.188254 | 1.198674 | 1.199382 | 1.19708 | 1.296532 |
| 3 | apply+mad | 19.837342 | 19.608640 | 20.059967 | 20.099953 | 20.05704 | 31.622045 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 0.288287 | 0.317768 | 0.3299942 | 0.3220715 | 0.3280665 | 0.450206 |
| 2 | rowMads2 | 0.340417 | 0.376679 | 0.3899930 | 0.3837420 | 0.3938805 | 0.521021 |
| 3 | apply+mad | 5.549613 | 6.125216 | 6.3968007 | 6.2457055 | 6.4645205 | 15.384339 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.00000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.180827 | 1.18539 | 1.181818 | 1.191481 | 1.200612 | 1.157295 |
| 3 | apply+mad | 19.250306 | 19.27575 | 19.384583 | 19.392295 | 19.704909 | 34.171777 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on integer+100x100 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() on integer+100x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 280.159 | 308.6925 | 319.2907 | 313.3360 | 322.6325 | 482.616 |
| 2 | rowMads | 288.287 | 317.7680 | 329.9942 | 322.0715 | 328.0665 | 450.206 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.0000 | 1.000000 | 1.000000 | 1.000000 | 1.0000000 |
| 2 | rowMads | 1.029012 | 1.0294 | 1.033523 | 1.027879 | 1.016843 | 0.9328452 |
Figure: Benchmarking of colMads() and rowMads() 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 5226745 279.2 8529671 455.6 8529671 455.6
Vcells 9683659 73.9 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5226721 279.2 8529671 455.6 8529671 455.6
Vcells 9688672 74.0 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on integer+1000x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 0.207184 | 0.2181840 | 0.2354643 | 0.2263615 | 0.2370005 | 0.344432 |
| 2 | colMads2 | 0.302347 | 0.3119515 | 0.3411153 | 0.3294780 | 0.3426210 | 0.533024 |
| 3 | apply+mad | 0.960231 | 1.0187745 | 1.0961040 | 1.0614555 | 1.1074695 | 1.758757 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | colMads2 | 1.459316 | 1.429763 | 1.448692 | 1.455539 | 1.445655 | 1.547545 |
| 3 | apply+mad | 4.634677 | 4.669336 | 4.655074 | 4.689205 | 4.672857 | 5.106253 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 0.214533 | 0.2274505 | 0.2468018 | 0.235193 | 0.263500 | 0.356902 |
| 2 | rowMads2 | 0.313418 | 0.3215165 | 0.3523157 | 0.341477 | 0.359623 | 0.557957 |
| 3 | apply+mad | 0.963733 | 1.0034030 | 1.0843441 | 1.045441 | 1.085357 | 1.718455 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.460931 | 1.413567 | 1.427525 | 1.451901 | 1.364793 | 1.563334 |
| 3 | apply+mad | 4.492237 | 4.411523 | 4.393582 | 4.445035 | 4.119000 | 4.814921 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on integer+1000x10 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() on integer+1000x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 207.184 | 218.1840 | 235.4643 | 226.3615 | 237.0005 | 344.432 |
| 2 | rowMads | 214.533 | 227.4505 | 246.8018 | 235.1930 | 263.5000 | 356.902 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.00000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads | 1.035471 | 1.042471 | 1.04815 | 1.039015 | 1.111812 | 1.036204 |
Figure: Benchmarking of colMads() and rowMads() 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 5226940 279.2 8529671 455.6 8529671 455.6
Vcells 9684074 73.9 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5226934 279.2 8529671 455.6 8529671 455.6
Vcells 9689117 74.0 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on integer+10x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 0.283539 | 0.3228190 | 0.3400543 | 0.3320985 | 0.352115 | 0.476548 |
| 2 | colMads2 | 0.372100 | 0.4230135 | 0.4478625 | 0.4414430 | 0.458013 | 0.642775 |
| 3 | apply+mad | 51.449480 | 56.7547215 | 61.5407343 | 57.3257015 | 58.492311 | 416.270212 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | colMads2 | 1.312342 | 1.310374 | 1.317032 | 1.329253 | 1.300748 | 1.348815 |
| 3 | apply+mad | 181.454685 | 175.809731 | 180.973257 | 172.616563 | 166.117067 | 873.511613 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 0.287380 | 0.3372805 | 0.3545651 | 0.3490950 | 0.3661025 | 0.509743 |
| 2 | rowMads2 | 0.376702 | 0.4239780 | 0.4518364 | 0.4408745 | 0.4601415 | 1.195377 |
| 3 | apply+mad | 53.628257 | 56.8922530 | 58.6950071 | 58.1868010 | 58.9637595 | 73.106085 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.00000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.310815 | 1.257049 | 1.27434 | 1.262907 | 1.256865 | 2.345058 |
| 3 | apply+mad | 186.610958 | 168.679343 | 165.54084 | 166.678987 | 161.058063 | 143.417536 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on integer+10x1000 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() 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 | colMads | 283.539 | 322.8190 | 340.0543 | 332.0985 | 352.1150 | 476.548 |
| 2 | rowMads | 287.380 | 337.2805 | 354.5651 | 349.0950 | 366.1025 | 509.743 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads | 1.013547 | 1.044798 | 1.042672 | 1.051179 | 1.039724 | 1.069657 |
Figure: Benchmarking of colMads() and rowMads() 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 5227144 279.2 8529671 455.6 8529671 455.6
Vcells 9685124 73.9 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5227138 279.2 8529671 455.6 8529671 455.6
Vcells 9735167 74.3 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on integer+100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 2.897207 | 3.174434 | 3.187247 | 3.189120 | 3.232470 | 3.555109 |
| 2 | colMads2 | 3.427215 | 3.786783 | 3.821480 | 3.807254 | 3.866204 | 4.937695 |
| 3 | apply+mad | 59.725124 | 63.099864 | 65.615657 | 64.223503 | 65.455270 | 76.280149 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.00000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | colMads2 | 1.182938 | 1.19290 | 1.198991 | 1.193826 | 1.196053 | 1.388901 |
| 3 | apply+mad | 20.614725 | 19.87752 | 20.586939 | 20.138315 | 20.249308 | 21.456487 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 2.869750 | 3.266530 | 3.348353 | 3.307662 | 3.382273 | 5.142912 |
| 2 | rowMads2 | 3.573523 | 3.840384 | 3.916673 | 3.932559 | 3.988018 | 4.423820 |
| 3 | apply+mad | 59.988647 | 63.974414 | 66.478240 | 64.797894 | 66.522702 | 89.255271 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.245238 | 1.175677 | 1.169731 | 1.188924 | 1.179094 | 0.860178 |
| 3 | apply+mad | 20.903789 | 19.584824 | 19.854012 | 19.590243 | 19.668049 | 17.355006 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on integer+100x1000 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() 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 | colMads | 2.897207 | 3.174434 | 3.187247 | 3.189120 | 3.232470 | 3.555109 |
| 2 | rowMads | 2.869750 | 3.266530 | 3.348353 | 3.307662 | 3.382273 | 5.142912 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.0000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads | 0.9905229 | 1.029012 | 1.050547 | 1.037171 | 1.046343 | 1.446626 |
Figure: Benchmarking of colMads() and rowMads() 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 5227350 279.2 8529671 455.6 8529671 455.6
Vcells 9685802 73.9 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5227344 279.2 8529671 455.6 8529671 455.6
Vcells 9735845 74.3 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on integer+1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 2.204126 | 2.521503 | 2.584057 | 2.536456 | 2.606876 | 3.609728 |
| 2 | colMads2 | 3.021677 | 3.420520 | 3.550127 | 3.441163 | 3.503141 | 12.308642 |
| 3 | apply+mad | 9.472349 | 10.531190 | 10.931617 | 10.619740 | 10.861638 | 21.006739 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | colMads2 | 1.370918 | 1.356540 | 1.373858 | 1.356681 | 1.343808 | 3.409853 |
| 3 | apply+mad | 4.297553 | 4.176553 | 4.230409 | 4.186841 | 4.166534 | 5.819480 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 2.281857 | 2.614507 | 2.662678 | 2.624229 | 2.664994 | 3.561178 |
| 2 | rowMads2 | 3.064162 | 3.471974 | 3.561733 | 3.499127 | 3.551153 | 4.948618 |
| 3 | apply+mad | 9.896669 | 10.569337 | 10.988476 | 10.634806 | 10.774353 | 20.211581 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.342837 | 1.327965 | 1.337651 | 1.333392 | 1.332518 | 1.389601 |
| 3 | apply+mad | 4.337112 | 4.042573 | 4.126851 | 4.052546 | 4.042918 | 5.675532 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on integer+1000x100 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() on integer+1000x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 2.204126 | 2.521503 | 2.584057 | 2.536456 | 2.606876 | 3.609728 |
| 2 | rowMads | 2.281857 | 2.614507 | 2.662678 | 2.624229 | 2.664994 | 3.561178 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.0000000 |
| 2 | rowMads | 1.035266 | 1.036884 | 1.030426 | 1.034604 | 1.022294 | 0.9865502 |
Figure: Benchmarking of colMads() and rowMads() 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 5227564 279.2 8529671 455.6 8529671 455.6
Vcells 9801659 74.8 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5227549 279.2 8529671 455.6 8529671 455.6
Vcells 9801787 74.8 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() 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 | colMads | 0.004581 | 0.0057935 | 0.0075455 | 0.0070780 | 0.0084730 | 0.023703 |
| 2 | colMads2 | 0.006717 | 0.0090925 | 0.0120188 | 0.0108080 | 0.0132355 | 0.043478 |
| 3 | apply+mad | 0.515738 | 0.5603135 | 0.6544891 | 0.6336655 | 0.7132445 | 1.163691 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.00000 | 1.000000 |
| 2 | colMads2 | 1.466274 | 1.569431 | 1.592853 | 1.526985 | 1.56208 | 1.834283 |
| 3 | apply+mad | 112.581969 | 96.714162 | 86.739221 | 89.526067 | 84.17851 | 49.094672 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() on double+10x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 0.004459 | 0.0060485 | 0.0077281 | 0.0075575 | 0.0085845 | 0.028484 |
| 2 | rowMads2 | 0.006522 | 0.0088780 | 0.0111525 | 0.0107755 | 0.0129125 | 0.027046 |
| 3 | apply+mad | 0.509955 | 0.5394140 | 0.6248800 | 0.6081865 | 0.6704965 | 1.007571 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.00000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.0000000 |
| 2 | rowMads2 | 1.46266 | 1.467802 | 1.443111 | 1.425802 | 1.504164 | 0.9495155 |
| 3 | apply+mad | 114.36533 | 89.181450 | 80.858372 | 80.474562 | 78.105481 | 35.3732271 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on double+10x10 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() on double+10x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 4.581 | 5.7935 | 7.54548 | 7.0780 | 8.4730 | 23.703 |
| 2 | rowMads | 4.459 | 6.0485 | 7.72808 | 7.5575 | 8.5845 | 28.484 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.0000000 | 1.000000 | 1.0000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads | 0.9733683 | 1.044015 | 1.0242 | 1.067745 | 1.013159 | 1.201704 |
Figure: Benchmarking of colMads() and rowMads() 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 5227760 279.2 8529671 455.6 8529671 455.6
Vcells 9802574 74.8 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5227754 279.2 8529671 455.6 8529671 455.6
Vcells 9812617 74.9 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on double+100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 0.365224 | 0.4220775 | 0.4516445 | 0.4607165 | 0.4767705 | 0.660302 |
| 2 | colMads2 | 0.396440 | 0.4572970 | 0.4835784 | 0.4871105 | 0.5106065 | 0.618663 |
| 3 | apply+mad | 5.544954 | 6.5021910 | 7.0000483 | 7.0005265 | 7.2346420 | 16.753403 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.0000000 |
| 2 | colMads2 | 1.085471 | 1.083443 | 1.070706 | 1.057289 | 1.070969 | 0.9369395 |
| 3 | apply+mad | 15.182337 | 15.405206 | 15.499023 | 15.194868 | 15.174265 | 25.3723342 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 0.394444 | 0.420360 | 0.4347465 | 0.4279820 | 0.4378975 | 0.573436 |
| 2 | rowMads2 | 0.427857 | 0.454976 | 0.4708328 | 0.4650805 | 0.4803030 | 0.611249 |
| 3 | apply+mad | 5.739571 | 6.473420 | 6.7375502 | 6.6279680 | 6.7521545 | 16.327633 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.084709 | 1.082348 | 1.083006 | 1.086682 | 1.096839 | 1.065941 |
| 3 | apply+mad | 14.551042 | 15.399705 | 15.497654 | 15.486558 | 15.419486 | 28.473331 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on double+100x100 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() 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 | |
|---|---|---|---|---|---|---|---|
| 2 | rowMads | 394.444 | 420.3600 | 434.7465 | 427.9820 | 437.8975 | 573.436 |
| 1 | colMads | 365.224 | 422.0775 | 451.6445 | 460.7165 | 476.7705 | 660.302 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 2 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 1 | colMads | 0.925921 | 1.004086 | 1.038869 | 1.076486 | 1.088772 | 1.151483 |
Figure: Benchmarking of colMads() and rowMads() 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 5227981 279.3 8529671 455.6 8529671 455.6
Vcells 9802734 74.8 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5227963 279.3 8529671 455.6 8529671 455.6
Vcells 9812757 74.9 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on double+1000x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 0.330291 | 0.3593555 | 0.3824990 | 0.3708205 | 0.3850245 | 0.546655 |
| 2 | colMads2 | 0.364039 | 0.3980275 | 0.4216585 | 0.4097920 | 0.4221325 | 0.629301 |
| 3 | apply+mad | 1.005330 | 1.1020030 | 1.1686173 | 1.1468255 | 1.1839375 | 1.779101 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | colMads2 | 1.102177 | 1.107615 | 1.102378 | 1.105095 | 1.096378 | 1.151185 |
| 3 | apply+mad | 3.043770 | 3.066610 | 3.055216 | 3.092670 | 3.074967 | 3.254523 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 0.327958 | 0.3568210 | 0.3809122 | 0.3698970 | 0.3919385 | 0.539816 |
| 2 | rowMads2 | 0.369916 | 0.4044725 | 0.4241505 | 0.4166235 | 0.4293825 | 0.642681 |
| 3 | apply+mad | 1.014903 | 1.0970550 | 1.1561683 | 1.1318380 | 1.1593105 | 1.784336 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.127937 | 1.133545 | 1.113513 | 1.126323 | 1.095535 | 1.190556 |
| 3 | apply+mad | 3.094613 | 3.074525 | 3.035262 | 3.059873 | 2.957889 | 3.305452 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on double+1000x10 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() 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 | |
|---|---|---|---|---|---|---|---|
| 2 | rowMads | 327.958 | 356.8210 | 380.9122 | 369.8970 | 391.9385 | 539.816 |
| 1 | colMads | 330.291 | 359.3555 | 382.4990 | 370.8205 | 385.0245 | 546.655 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 2 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.0000000 | 1.000000 |
| 1 | colMads | 1.007114 | 1.007103 | 1.004166 | 1.002497 | 0.9823595 | 1.012669 |
Figure: Benchmarking of colMads() and rowMads() 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 5228182 279.3 8529671 455.6 8529671 455.6
Vcells 9803801 74.8 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5228176 279.3 8529671 455.6 8529671 455.6
Vcells 9813844 74.9 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on double+10x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 0.363888 | 0.4209730 | 0.4391614 | 0.432610 | 0.4420745 | 0.604191 |
| 2 | colMads2 | 0.428136 | 0.4918125 | 0.5208620 | 0.510509 | 0.5270265 | 0.728955 |
| 3 | apply+mad | 52.467798 | 57.3641435 | 59.0773159 | 58.473685 | 59.5960105 | 73.978570 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.00000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | colMads2 | 1.17656 | 1.168276 | 1.186038 | 1.180068 | 1.192167 | 1.206498 |
| 3 | apply+mad | 144.18667 | 136.265612 | 134.523007 | 135.164895 | 134.809880 | 122.442357 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 0.359739 | 0.4220620 | 0.4412593 | 0.4350155 | 0.448377 | 0.686357 |
| 2 | rowMads2 | 0.424041 | 0.4981685 | 0.5205947 | 0.5077900 | 0.518982 | 1.316516 |
| 3 | apply+mad | 53.433170 | 58.1170070 | 59.8168121 | 59.1039425 | 60.135291 | 82.647628 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.178746 | 1.180321 | 1.179793 | 1.167292 | 1.157468 | 1.918121 |
| 3 | apply+mad | 148.533159 | 137.697796 | 135.559336 | 135.866291 | 134.117698 | 120.414927 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on double+10x1000 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() 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 | colMads | 363.888 | 420.973 | 439.1614 | 432.6100 | 442.0745 | 604.191 |
| 2 | rowMads | 359.739 | 422.062 | 441.2593 | 435.0155 | 448.3770 | 686.357 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.0000000 | 1.000000 | 1.000000 | 1.00000 | 1.000000 | 1.000000 |
| 2 | rowMads | 0.9885981 | 1.002587 | 1.004777 | 1.00556 | 1.014257 | 1.135993 |
Figure: Benchmarking of colMads() and rowMads() 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 5228386 279.3 8529671 455.6 8529671 455.6
Vcells 9803936 74.8 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5228380 279.3 8529671 455.6 8529671 455.6
Vcells 9903979 75.6 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on double+100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 3.956343 | 4.294430 | 4.376776 | 4.398336 | 4.444208 | 4.714913 |
| 2 | colMads2 | 4.097677 | 4.488697 | 4.619376 | 4.609297 | 4.680533 | 6.337030 |
| 3 | apply+mad | 58.779160 | 64.558275 | 70.440677 | 65.487746 | 66.503474 | 426.564833 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.00000 |
| 2 | colMads2 | 1.035723 | 1.045237 | 1.055429 | 1.047964 | 1.053176 | 1.34404 |
| 3 | apply+mad | 14.856942 | 15.033024 | 16.094192 | 14.889209 | 14.964080 | 90.47141 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 3.937613 | 4.371682 | 4.477499 | 4.430325 | 4.532702 | 5.948798 |
| 2 | rowMads2 | 4.130020 | 4.604520 | 4.695753 | 4.662670 | 4.819568 | 6.432425 |
| 3 | apply+mad | 61.579233 | 64.934735 | 67.588507 | 65.759079 | 66.908979 | 91.575860 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.048864 | 1.053261 | 1.048745 | 1.052444 | 1.063288 | 1.081298 |
| 3 | apply+mad | 15.638722 | 14.853492 | 15.095149 | 14.842947 | 14.761389 | 15.394011 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on double+100x1000 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() on double+100x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 3.956343 | 4.294430 | 4.376776 | 4.398336 | 4.444208 | 4.714913 |
| 2 | rowMads | 3.937613 | 4.371682 | 4.477499 | 4.430325 | 4.532702 | 5.948798 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.0000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads | 0.9952658 | 1.017989 | 1.023013 | 1.007273 | 1.019912 | 1.261698 |
Figure: Benchmarking of colMads() and rowMads() 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 5228592 279.3 8529671 455.6 8529671 455.6
Vcells 9805233 74.9 31876688 243.2 60562128 462.1
> colStats <- microbenchmark(colMads = colMads(X, na.rm = FALSE), colMads2 = colMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 2L, FUN = mad, na.rm = FALSE), unit = "ms")
> X <- t(X)
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 5228586 279.3 8529671 455.6 8529671 455.6
Vcells 9905276 75.6 31876688 243.2 60562128 462.1
> rowStats <- microbenchmark(rowMads = rowMads(X, na.rm = FALSE), rowMads2 = rowMads2(X, na.rm = FALSE),
+ `apply+mad` = apply(X, MARGIN = 1L, FUN = mad, na.rm = FALSE), unit = "ms")
Table: Benchmarking of colMads(), colMads2() and apply+mad() on double+1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 3.442158 | 3.920787 | 4.025222 | 3.975932 | 4.055993 | 5.551028 |
| 2 | colMads2 | 3.849091 | 4.211763 | 4.293943 | 4.283513 | 4.360643 | 4.712523 |
| 3 | apply+mad | 10.123288 | 11.203198 | 11.599529 | 11.338291 | 11.476510 | 20.073749 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | colMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | colMads2 | 1.118220 | 1.074214 | 1.066759 | 1.077361 | 1.075111 | 0.848946 |
| 3 | apply+mad | 2.940971 | 2.857385 | 2.881712 | 2.851732 | 2.829519 | 3.616222 |
Table: Benchmarking of rowMads(), rowMads2() and apply+mad() 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 | rowMads | 3.504217 | 3.903222 | 3.979249 | 3.963858 | 4.030054 | 4.896801 |
| 2 | rowMads2 | 3.837550 | 4.190633 | 4.396986 | 4.276613 | 4.335335 | 11.818425 |
| 3 | apply+mad | 9.987954 | 11.275250 | 11.559008 | 11.384596 | 11.506334 | 20.089477 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 1 | rowMads | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 2 | rowMads2 | 1.095123 | 1.073634 | 1.104979 | 1.078902 | 1.075751 | 2.413499 |
| 3 | apply+mad | 2.850267 | 2.888703 | 2.904822 | 2.872100 | 2.855132 | 4.102572 |
Figure: Benchmarking of colMads(), colMads2() and apply+mad() on double+1000x100 data as well as rowMads(), rowMads2() and apply+mad() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.
Table: Benchmarking of colMads() and rowMads() 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 | rowMads | 3.504217 | 3.903222 | 3.979249 | 3.963858 | 4.030054 | 4.896801 |
| 1 | colMads | 3.442158 | 3.920787 | 4.025222 | 3.975932 | 4.055993 | 5.551028 |
| expr | min | lq | mean | median | uq | max | |
|---|---|---|---|---|---|---|---|
| 2 | rowMads | 1.0000000 | 1.0000 | 1.000000 | 1.000000 | 1.000000 | 1.000000 |
| 1 | colMads | 0.9822902 | 1.0045 | 1.011553 | 1.003046 | 1.006436 | 1.133603 |
Figure: Benchmarking of colMads() and rowMads() on double+1000x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.
Appendix
Session information
R version 4.1.1 Patched (2021-08-10 r80727)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.5 LTS
Matrix products: default
BLAS: /home/hb/software/R-devel/R-4-1-branch/lib/R/lib/libRblas.so
LAPACK: /home/hb/software/R-devel/R-4-1-branch/lib/R/lib/libRlapack.so
locale:
[1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
[3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
[5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
[7] LC_PAPER=en_US.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] microbenchmark_1.4-7 matrixStats_0.60.1 ggplot2_3.3.5
[4] knitr_1.33 R.devices_2.17.0 R.utils_2.10.1
[7] R.oo_1.24.0 R.methodsS3_1.8.1-9001 history_0.0.1-9000
loaded via a namespace (and not attached):
[1] Biobase_2.52.0 httr_1.4.2 splines_4.1.1
[4] bit64_4.0.5 network_1.17.1 assertthat_0.2.1
[7] highr_0.9 stats4_4.1.1 blob_1.2.2
[10] GenomeInfoDbData_1.2.6 robustbase_0.93-8 pillar_1.6.2
[13] RSQLite_2.2.8 lattice_0.20-44 glue_1.4.2
[16] digest_0.6.27 XVector_0.32.0 colorspace_2.0-2
[19] Matrix_1.3-4 XML_3.99-0.7 pkgconfig_2.0.3
[22] zlibbioc_1.38.0 genefilter_1.74.0 purrr_0.3.4
[25] ergm_4.1.2 xtable_1.8-4 scales_1.1.1
[28] tibble_3.1.4 annotate_1.70.0 KEGGREST_1.32.0
[31] farver_2.1.0 generics_0.1.0 IRanges_2.26.0
[34] ellipsis_0.3.2 cachem_1.0.6 withr_2.4.2
[37] BiocGenerics_0.38.0 mime_0.11 survival_3.2-13
[40] magrittr_2.0.1 crayon_1.4.1 statnet.common_4.5.0
[43] memoise_2.0.0 laeken_0.5.1 fansi_0.5.0
[46] R.cache_0.15.0 MASS_7.3-54 R.rsp_0.44.0
[49] progressr_0.8.0 tools_4.1.1 lifecycle_1.0.0
[52] S4Vectors_0.30.0 trust_0.1-8 munsell_0.5.0
[55] tabby_0.0.1-9001 AnnotationDbi_1.54.1 Biostrings_2.60.2
[58] compiler_4.1.1 GenomeInfoDb_1.28.1 rlang_0.4.11
[61] grid_4.1.1 RCurl_1.98-1.4 cwhmisc_6.6
[64] rappdirs_0.3.3 startup_0.15.0 labeling_0.4.2
[67] bitops_1.0-7 base64enc_0.1-3 boot_1.3-28
[70] gtable_0.3.0 DBI_1.1.1 markdown_1.1
[73] R6_2.5.1 lpSolveAPI_5.5.2.0-17.7 rle_0.9.2
[76] dplyr_1.0.7 fastmap_1.1.0 bit_4.0.4
[79] utf8_1.2.2 parallel_4.1.1 Rcpp_1.0.7
[82] vctrs_0.3.8 png_0.1-7 DEoptimR_1.0-9
[85] tidyselect_1.1.1 xfun_0.25 coda_0.19-4
Total processing time was 1.48 mins.
Reproducibility
To reproduce this report, do:
html <- matrixStats:::benchmark('colMads')
Copyright Henrik Bengtsson. Last updated on 2021-08-25 18:58:11 (+0200 UTC). Powered by RSP.