Skip to content

colRowWeightedMeans_subset

Jump to bottom Edit New page
Henrik Bengtsson edited this page Sep 11, 2019 · 3 revisions

matrixStats: Benchmark report

colWeightedMeans() and rowWeightedMeans() benchmarks on subsetted computation

This report benchmark the performance of colWeightedMeans() and rowWeightedMeans() on subsetted computation.

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 = "double")

Results

10x10 matrix

> X <- data[["10x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> w <- runif(nrow(X))
> w_S <- w[rows]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3220830 172.1    5709258 305.0  5709258 305.0
Vcells 6731746  51.4   22345847 170.5 56666022 432.4
> colStats <- microbenchmark(colWeightedMeans_X_w_S = colWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `colWeightedMeans(X, w, rows, cols)` = colWeightedMeans(X, w = w, rows = rows, cols = cols, na.rm = FALSE), 
+     `colWeightedMeans(X[rows, cols], w[rows])` = colWeightedMeans(X[rows, cols], w = w[rows], na.rm = FALSE), 
+     unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3219991 172.0    5709258 305.0  5709258 305.0
Vcells 6729258  51.4   22345847 170.5 56666022 432.4
> rowStats <- microbenchmark(rowWeightedMeans_X_w_S = rowWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `rowWeightedMeans(X, w, cols, rows)` = rowWeightedMeans(X, w = w, rows = cols, cols = rows, na.rm = FALSE), 
+     `rowWeightedMeans(X[cols, rows], w[rows])` = rowWeightedMeans(X[cols, rows], w = w[rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 10x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

expr min lq mean median uq max
1 colWeightedMeans_X_w_S 0.008422 0.0088190 0.0090774 0.0090070 0.0092395 0.016219
3 colWeightedMeans(X[rows, cols], w[rows]) 0.009737 0.0102495 0.0104549 0.0104245 0.0105650 0.014009
2 colWeightedMeans(X, w, rows, cols) 0.010534 0.0111485 0.0145679 0.0113280 0.0115600 0.325739
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.00000 1.0000000
3 colWeightedMeans(X[rows, cols], w[rows]) 1.156139 1.162207 1.151757 1.157378 1.14346 0.8637401
2 colWeightedMeans(X, w, rows, cols) 1.250772 1.264146 1.604858 1.257688 1.25115 20.0837906

Table: Benchmarking of rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on 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 rowWeightedMeans_X_w_S 0.010462 0.0106835 0.0109957 0.0108735 0.011067 0.015903
3 rowWeightedMeans(X[cols, rows], w[rows]) 0.011303 0.0119175 0.0123037 0.0121050 0.012298 0.021273
2 rowWeightedMeans(X, w, cols, rows) 0.012672 0.0129740 0.0166216 0.0131220 0.013385 0.345761
expr min lq mean median uq max
1 rowWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
3 rowWeightedMeans(X[cols, rows], w[rows]) 1.080386 1.115505 1.118956 1.113257 1.111232 1.337672
2 rowWeightedMeans(X, w, cols, rows) 1.211241 1.214396 1.511647 1.206787 1.209451 21.741873

Figure: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 10x10 data as well as rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 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 colWeightedMeans_X_w_S 8.422 8.8190 9.07736 9.0070 9.2395 16.219
2 rowWeightedMeans_X_w_S 10.462 10.6835 10.99570 10.8735 11.0670 15.903
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.0000000
2 rowWeightedMeans_X_w_S 1.242223 1.211418 1.211332 1.207228 1.197792 0.9805167

Figure: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 10x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x100 matrix

> X <- data[["100x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> w <- runif(nrow(X))
> w_S <- w[rows]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3218973 172.0    5709258 305.0  5709258 305.0
Vcells 6399938  48.9   22345847 170.5 56666022 432.4
> colStats <- microbenchmark(colWeightedMeans_X_w_S = colWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `colWeightedMeans(X, w, rows, cols)` = colWeightedMeans(X, w = w, rows = rows, cols = cols, na.rm = FALSE), 
+     `colWeightedMeans(X[rows, cols], w[rows])` = colWeightedMeans(X[rows, cols], w = w[rows], na.rm = FALSE), 
+     unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3218964  172    5709258 305.0  5709258 305.0
Vcells 6410016   49   22345847 170.5 56666022 432.4
> rowStats <- microbenchmark(rowWeightedMeans_X_w_S = rowWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `rowWeightedMeans(X, w, cols, rows)` = rowWeightedMeans(X, w = w, rows = cols, cols = rows, na.rm = FALSE), 
+     `rowWeightedMeans(X[cols, rows], w[rows])` = rowWeightedMeans(X[cols, rows], w = w[rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

expr min lq mean median uq max
1 colWeightedMeans_X_w_S 0.018505 0.0191635 0.0200976 0.0197605 0.0206115 0.026853
3 colWeightedMeans(X[rows, cols], w[rows]) 0.034474 0.0357510 0.0375220 0.0362285 0.0375605 0.073949
2 colWeightedMeans(X, w, rows, cols) 0.042417 0.0432050 0.0457495 0.0438460 0.0459230 0.113261
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
3 colWeightedMeans(X[rows, cols], w[rows]) 1.862956 1.865578 1.866988 1.833380 1.822308 2.753845
2 colWeightedMeans(X, w, rows, cols) 2.292191 2.254546 2.276367 2.218871 2.228028 4.217816

Table: Benchmarking of rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on 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 rowWeightedMeans_X_w_S 0.055723 0.056827 0.0575488 0.0572280 0.0579205 0.067981
3 rowWeightedMeans(X[cols, rows], w[rows]) 0.065264 0.065902 0.0668488 0.0664445 0.0673160 0.076393
2 rowWeightedMeans(X, w, cols, rows) 0.072875 0.073695 0.0756670 0.0745210 0.0757430 0.138304
expr min lq mean median uq max
1 rowWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
3 rowWeightedMeans(X[cols, rows], w[rows]) 1.171222 1.159695 1.161601 1.161049 1.162214 1.123741
2 rowWeightedMeans(X, w, cols, rows) 1.307808 1.296831 1.314832 1.302177 1.307706 2.034451

Figure: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 100x100 data as well as rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 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 colWeightedMeans_X_w_S 18.505 19.1635 20.09761 19.7605 20.6115 26.853
2 rowWeightedMeans_X_w_S 55.723 56.8270 57.54882 57.2280 57.9205 67.981
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.00000 1.000000 1.000000 1.000000 1.000000 1.000000
2 rowWeightedMeans_X_w_S 3.01124 2.965377 2.863466 2.896081 2.810106 2.531598

Figure: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 100x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x10 matrix

> X <- data[["1000x10"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> w <- runif(nrow(X))
> w_S <- w[rows]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3219730 172.0    5709258 305.0  5709258 305.0
Vcells 6405864  48.9   22345847 170.5 56666022 432.4
> colStats <- microbenchmark(colWeightedMeans_X_w_S = colWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `colWeightedMeans(X, w, rows, cols)` = colWeightedMeans(X, w = w, rows = rows, cols = cols, na.rm = FALSE), 
+     `colWeightedMeans(X[rows, cols], w[rows])` = colWeightedMeans(X[rows, cols], w = w[rows], na.rm = FALSE), 
+     unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3219721  172    5709258 305.0  5709258 305.0
Vcells 6415942   49   22345847 170.5 56666022 432.4
> rowStats <- microbenchmark(rowWeightedMeans_X_w_S = rowWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `rowWeightedMeans(X, w, cols, rows)` = rowWeightedMeans(X, w = w, rows = cols, cols = rows, na.rm = FALSE), 
+     `rowWeightedMeans(X[cols, rows], w[rows])` = rowWeightedMeans(X[cols, rows], w = w[rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 1000x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

expr min lq mean median uq max
1 colWeightedMeans_X_w_S 0.026795 0.0274550 0.0287832 0.0277690 0.029024 0.058835
3 colWeightedMeans(X[rows, cols], w[rows]) 0.044465 0.0453660 0.0467978 0.0460920 0.047612 0.060079
2 colWeightedMeans(X, w, rows, cols) 0.054934 0.0558125 0.0576411 0.0563185 0.057492 0.137369
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
3 colWeightedMeans(X[rows, cols], w[rows]) 1.659451 1.652377 1.625871 1.659836 1.640435 1.021144
2 colWeightedMeans(X, w, rows, cols) 2.050159 2.032872 2.002590 2.028107 1.980843 2.334818

Table: Benchmarking of rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on 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 rowWeightedMeans_X_w_S 0.065006 0.0658005 0.0664799 0.066194 0.0666275 0.071336
3 rowWeightedMeans(X[cols, rows], w[rows]) 0.077246 0.0780405 0.0795255 0.078481 0.0792495 0.150825
2 rowWeightedMeans(X, w, cols, rows) 0.089107 0.0902450 0.0916613 0.090808 0.0916245 0.121058
expr min lq mean median uq max
1 rowWeightedMeans_X_w_S 1.00000 1.000000 1.000000 1.000000 1.000000 1.000000
3 rowWeightedMeans(X[cols, rows], w[rows]) 1.18829 1.186017 1.196234 1.185621 1.189441 2.114290
2 rowWeightedMeans(X, w, cols, rows) 1.37075 1.371494 1.378782 1.371846 1.375175 1.697011

Figure: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 1000x10 data as well as rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 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 colWeightedMeans_X_w_S 26.795 27.4550 28.78325 27.769 29.0240 58.835
2 rowWeightedMeans_X_w_S 65.006 65.8005 66.47990 66.194 66.6275 71.336
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.00000 1.000000 1.000000 1.000000 1.0000 1.000000
2 rowWeightedMeans_X_w_S 2.42605 2.396667 2.309673 2.383737 2.2956 1.212476

Figure: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 1000x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

10x1000 matrix

> X <- data[["10x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> w <- runif(nrow(X))
> w_S <- w[rows]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3219930 172.0    5709258 305.0  5709258 305.0
Vcells 6405137  48.9   22345847 170.5 56666022 432.4
> colStats <- microbenchmark(colWeightedMeans_X_w_S = colWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `colWeightedMeans(X, w, rows, cols)` = colWeightedMeans(X, w = w, rows = rows, cols = cols, na.rm = FALSE), 
+     `colWeightedMeans(X[rows, cols], w[rows])` = colWeightedMeans(X[rows, cols], w = w[rows], na.rm = FALSE), 
+     unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3219921  172    5709258 305.0  5709258 305.0
Vcells 6415215   49   22345847 170.5 56666022 432.4
> rowStats <- microbenchmark(rowWeightedMeans_X_w_S = rowWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `rowWeightedMeans(X, w, cols, rows)` = rowWeightedMeans(X, w = w, rows = cols, cols = rows, na.rm = FALSE), 
+     `rowWeightedMeans(X[cols, rows], w[rows])` = rowWeightedMeans(X[cols, rows], w = w[rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 10x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

expr min lq mean median uq max
1 colWeightedMeans_X_w_S 0.018081 0.0188180 0.0199830 0.0191135 0.019512 0.074175
3 colWeightedMeans(X[rows, cols], w[rows]) 0.028089 0.0289945 0.0294583 0.0292890 0.029721 0.032773
2 colWeightedMeans(X, w, rows, cols) 0.036647 0.0375030 0.0386619 0.0380405 0.038872 0.063118
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.0000000
3 colWeightedMeans(X[rows, cols], w[rows]) 1.553509 1.540785 1.474167 1.532372 1.523216 0.4418335
2 colWeightedMeans(X, w, rows, cols) 2.026824 1.992932 1.934739 1.990242 1.992210 0.8509336

Table: Benchmarking of rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on 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 rowWeightedMeans_X_w_S 0.054120 0.0549015 0.0555479 0.0551930 0.0555910 0.063898
3 rowWeightedMeans(X[cols, rows], w[rows]) 0.063296 0.0639285 0.0655693 0.0642665 0.0646320 0.154854
2 rowWeightedMeans(X, w, cols, rows) 0.070700 0.0717700 0.0726398 0.0721455 0.0725115 0.101257
expr min lq mean median uq max
1 rowWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
3 rowWeightedMeans(X[cols, rows], w[rows]) 1.169549 1.164422 1.180410 1.164396 1.162634 2.423456
2 rowWeightedMeans(X, w, cols, rows) 1.306356 1.307250 1.307696 1.307149 1.304375 1.584666

Figure: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 10x1000 data as well as rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 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 colWeightedMeans_X_w_S 18.081 18.8180 19.98299 19.1135 19.512 74.175
2 rowWeightedMeans_X_w_S 54.120 54.9015 55.54790 55.1930 55.591 63.898
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.0000000
2 rowWeightedMeans_X_w_S 2.993197 2.917499 2.779759 2.887645 2.849067 0.8614493

Figure: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 10x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x1000 matrix

> X <- data[["100x1000"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> w <- runif(nrow(X))
> w_S <- w[rows]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3220138 172.0    5709258 305.0  5709258 305.0
Vcells 6450084  49.3   22345847 170.5 56666022 432.4
> colStats <- microbenchmark(colWeightedMeans_X_w_S = colWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `colWeightedMeans(X, w, rows, cols)` = colWeightedMeans(X, w = w, rows = rows, cols = cols, na.rm = FALSE), 
+     `colWeightedMeans(X[rows, cols], w[rows])` = colWeightedMeans(X[rows, cols], w = w[rows], na.rm = FALSE), 
+     unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3220129  172    5709258 305.0  5709258 305.0
Vcells 6550162   50   22345847 170.5 56666022 432.4
> rowStats <- microbenchmark(rowWeightedMeans_X_w_S = rowWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `rowWeightedMeans(X, w, cols, rows)` = rowWeightedMeans(X, w = w, rows = cols, cols = rows, na.rm = FALSE), 
+     `rowWeightedMeans(X[cols, rows], w[rows])` = rowWeightedMeans(X[cols, rows], w = w[rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

expr min lq mean median uq max
1 colWeightedMeans_X_w_S 0.086715 0.0936375 0.1324199 0.1033760 0.1913845 0.216416
3 colWeightedMeans(X[rows, cols], w[rows]) 0.236394 0.2543960 0.3355398 0.2859300 0.4473855 0.478072
2 colWeightedMeans(X, w, rows, cols) 0.296554 0.3096775 0.4974049 0.3571465 0.6114485 6.401086
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
3 colWeightedMeans(X[rows, cols], w[rows]) 2.726103 2.716817 2.533907 2.765922 2.337627 2.209042
2 colWeightedMeans(X, w, rows, cols) 3.419870 3.307195 3.756269 3.454830 3.194869 29.577693

Table: Benchmarking of rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on 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 rowWeightedMeans_X_w_S 0.448142 0.4494210 0.4642071 0.4520655 0.4597655 0.585779
3 rowWeightedMeans(X[cols, rows], w[rows]) 0.517342 0.5194130 0.5953113 0.5225275 0.5274235 6.720449
2 rowWeightedMeans(X, w, cols, rows) 0.586399 0.5900345 0.6102561 0.5932900 0.6014325 0.749185
expr min lq mean median uq max
1 rowWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
3 rowWeightedMeans(X[cols, rows], w[rows]) 1.154415 1.155738 1.282426 1.155867 1.147158 11.472670
2 rowWeightedMeans(X, w, cols, rows) 1.308512 1.312877 1.314620 1.312398 1.308129 1.278955

Figure: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 100x1000 data as well as rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 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 colWeightedMeans_X_w_S 86.715 93.6375 132.4199 103.3760 191.3845 216.416
2 rowWeightedMeans_X_w_S 448.142 449.4210 464.2071 452.0655 459.7655 585.779
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
2 rowWeightedMeans_X_w_S 5.167987 4.799583 3.505568 4.373022 2.402313 2.706727

Figure: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 100x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x100 matrix

> X <- data[["1000x100"]]
> rows <- sample.int(nrow(X), size = nrow(X) * 0.7)
> cols <- sample.int(ncol(X), size = ncol(X) * 0.7)
> X_S <- X[rows, cols]
> w <- runif(nrow(X))
> w_S <- w[rows]
> gc()
          used  (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3220345 172.0    5709258 305.0  5709258 305.0
Vcells 6452397  49.3   22345847 170.5 56666022 432.4
> colStats <- microbenchmark(colWeightedMeans_X_w_S = colWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `colWeightedMeans(X, w, rows, cols)` = colWeightedMeans(X, w = w, rows = rows, cols = cols, na.rm = FALSE), 
+     `colWeightedMeans(X[rows, cols], w[rows])` = colWeightedMeans(X[rows, cols], w = w[rows], na.rm = FALSE), 
+     unit = "ms")
> X <- t(X)
> X_S <- t(X_S)
> gc()
          used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells 3220336  172    5709258 305.0  5709258 305.0
Vcells 6552475   50   22345847 170.5 56666022 432.4
> rowStats <- microbenchmark(rowWeightedMeans_X_w_S = rowWeightedMeans(X_S, w = w_S, na.rm = FALSE), 
+     `rowWeightedMeans(X, w, cols, rows)` = rowWeightedMeans(X, w = w, rows = cols, cols = rows, na.rm = FALSE), 
+     `rowWeightedMeans(X[cols, rows], w[rows])` = rowWeightedMeans(X[cols, rows], w = w[rows], na.rm = FALSE), 
+     unit = "ms")

Table: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

expr min lq mean median uq max
1 colWeightedMeans_X_w_S 0.104403 0.1366315 0.1931331 0.2080620 0.219613 0.327439
3 colWeightedMeans(X[rows, cols], w[rows]) 0.176720 0.2192125 0.4169919 0.3766440 0.390549 9.763244
2 colWeightedMeans(X, w, rows, cols) 0.238867 0.3015410 0.4855723 0.5448855 0.564166 0.682642
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
3 colWeightedMeans(X[rows, cols], w[rows]) 1.692672 1.604407 2.159091 1.810249 1.778351 29.816986
2 colWeightedMeans(X, w, rows, cols) 2.287932 2.206965 2.514185 2.618861 2.568910 2.084791

Table: Benchmarking of rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on 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 rowWeightedMeans_X_w_S 0.457238 0.4592535 0.4712499 0.4606355 0.4644000 0.597803
3 rowWeightedMeans(X[cols, rows], w[rows]) 0.533144 0.5368280 0.5588002 0.5394435 0.5507575 0.745953
2 rowWeightedMeans(X, w, cols, rows) 0.598959 0.6122735 0.7004234 0.6154305 0.6302045 6.803909
expr min lq mean median uq max
1 rowWeightedMeans_X_w_S 1.00000 1.000000 1.000000 1.000000 1.000000 1.000000
3 rowWeightedMeans(X[cols, rows], w[rows]) 1.16601 1.168914 1.185783 1.171085 1.185955 1.247824
2 rowWeightedMeans(X, w, cols, rows) 1.30995 1.333193 1.486310 1.336047 1.357030 11.381524

Figure: Benchmarking of colWeightedMeans_X_w_S(), colWeightedMeans(X, w, rows, cols)() and colWeightedMeans(X[rows, cols], w[rows])() on 1000x100 data as well as rowWeightedMeans_X_w_S(), rowWeightedMeans(X, w, cols, rows)() and rowWeightedMeans(X[cols, rows], w[rows])() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 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 colWeightedMeans_X_w_S 104.403 136.6315 193.1331 208.0620 219.613 327.439
2 rowWeightedMeans_X_w_S 457.238 459.2535 471.2499 460.6355 464.400 597.803
expr min lq mean median uq max
1 colWeightedMeans_X_w_S 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
2 rowWeightedMeans_X_w_S 4.379549 3.361256 2.440027 2.213934 2.114629 1.825693

Figure: Benchmarking of colWeightedMeans_X_w_S() and rowWeightedMeans_X_w_S() on 1000x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

Appendix

Session information

R version 3.6.1 Patched (2019-08-27 r77078)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.3 LTS

Matrix products: default
BLAS:   /home/hb/software/R-devel/R-3-6-branch/lib/R/lib/libRblas.so
LAPACK: /home/hb/software/R-devel/R-3-6-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_1.4-6    matrixStats_0.55.0-9000 ggplot2_3.2.1          
[4] knitr_1.24              R.devices_2.16.0        R.utils_2.9.0          
[7] R.oo_1.22.0             R.methodsS3_1.7.1       history_0.0.0-9002     

loaded via a namespace (and not attached):
 [1] Biobase_2.45.0       bit64_0.9-7          splines_3.6.1       
 [4] network_1.15         assertthat_0.2.1     highr_0.8           
 [7] stats4_3.6.1         blob_1.2.0           robustbase_0.93-5   
[10] pillar_1.4.2         RSQLite_2.1.2        backports_1.1.4     
[13] lattice_0.20-38      glue_1.3.1           digest_0.6.20       
[16] colorspace_1.4-1     sandwich_2.5-1       Matrix_1.2-17       
[19] XML_3.98-1.20        lpSolve_5.6.13.3     pkgconfig_2.0.2     
[22] genefilter_1.66.0    purrr_0.3.2          ergm_3.10.4         
[25] xtable_1.8-4         mvtnorm_1.0-11       scales_1.0.0        
[28] tibble_2.1.3         annotate_1.62.0      IRanges_2.18.2      
[31] TH.data_1.0-10       withr_2.1.2          BiocGenerics_0.30.0 
[34] lazyeval_0.2.2       mime_0.7             survival_2.44-1.1   
[37] magrittr_1.5         crayon_1.3.4         statnet.common_4.3.0
[40] memoise_1.1.0        laeken_0.5.0         R.cache_0.13.0      
[43] MASS_7.3-51.4        R.rsp_0.43.1         tools_3.6.1         
[46] multcomp_1.4-10      S4Vectors_0.22.1     trust_0.1-7         
[49] munsell_0.5.0        AnnotationDbi_1.46.1 compiler_3.6.1      
[52] rlang_0.4.0          grid_3.6.1           RCurl_1.95-4.12     
[55] cwhmisc_6.6          rappdirs_0.3.1       labeling_0.3        
[58] bitops_1.0-6         base64enc_0.1-3      boot_1.3-23         
[61] gtable_0.3.0         codetools_0.2-16     DBI_1.0.0           
[64] markdown_1.1         R6_2.4.0             zoo_1.8-6           
[67] dplyr_0.8.3          bit_1.1-14           zeallot_0.1.0       
[70] parallel_3.6.1       Rcpp_1.0.2           vctrs_0.2.0         
[73] DEoptimR_1.0-8       tidyselect_0.2.5     xfun_0.9            
[76] coda_0.19-3

Total processing time was 12.21 secs.

Reproducibility

To reproduce this report, do:

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

Copyright Dongcan Jiang. Last updated on 2019-09-10 20:55:21 (-0700 UTC). Powered by RSP.

<script> var link = document.createElement('link'); link.rel = 'icon'; link.href = "data:image/png;base64,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" document.getElementsByTagName('head')[0].appendChild(link); </script>
Add a custom footer

Benchmark reports

Clone this wiki locally