nodist/timings.md
In shabbychef/fromo: Fast Robust Moments
fromo timings
To prevent performance regressions, compare them here, including benchmarks
against other packages:
library(fromo)
library(RollingWindow)
library(roll)
library(RcppRoll)
library(microbenchmark)
library(moments)
library(RcppParallel)
# keep this constant for comparison
setThreadOptions(numThreads = 2)
print(Sys.info())
## sysname release version
## "Linux" "4.4.0-137-generic" "#163-Ubuntu SMP Mon Sep 24 13:14:43 UTC 2018"
## nodename machine login
## "2d2df85a7b70" "x86_64" "unknown"
## user effective_user
## "docker" "docker"
print(sessionInfo())
## R version 3.5.2 (2018-12-20)
## Platform: x86_64-pc-linux-gnu (64-bit)
## Running under: Debian GNU/Linux buster/sid
##
## Matrix products: default
## BLAS: /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.8.0
## LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.8.0
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8 LC_MONETARY=en_US.UTF-8
## [6] LC_MESSAGES=en_US.UTF-8 LC_PAPER=en_US.UTF-8 LC_NAME=C LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## attached base packages:
## [1] utils methods base
##
## other attached packages:
## [1] RcppParallel_4.4.2 RcppRoll_0.3.0 roll_1.1.2 RollingWindow_0.2 microbenchmark_1.4-6 moments_0.14 dplyr_0.7.8
## [8] ggplot2_3.1.0 knitr_1.21 fromo_0.1.3.6796
##
## loaded via a namespace (and not attached):
## [1] Rcpp_1.0.0 bindr_0.1.1 magrittr_1.5 grDevices_3.5.2 tidyselect_0.2.5 munsell_0.5.0 colorspace_1.3-2 R6_2.3.0 rlang_0.3.0.1
## [10] stringr_1.3.1 plyr_1.8.4 tools_3.5.2 grid_3.5.2 gtable_0.2.0 xfun_0.4 withr_2.1.2 stats_3.5.2 lazyeval_0.2.1
## [19] assertthat_0.2.0 tibble_1.4.2 crayon_1.3.4 bindrcpp_0.2.2 formatR_1.5 purrr_0.2.5 graphics_3.5.2 glue_1.3.0 evaluate_0.12
## [28] stringi_1.2.4 compiler_3.5.2 pillar_1.3.1 scales_1.0.0 pkgconfig_2.0.2
set.seed(12345)
x <- rnorm(1e+05)
xpos <- runif(length(x)) + 1
xm <- matrix(x, ncol = 1)
xmps <- matrix(xpos, ncol = 1)
w <- runif(length(x))
dumbk <- function(x) {
c(kurtosis(x) - 3, skewness(x), sd(x), mean(x),
length(x))
}
checkit <- microbenchmark(sum(x), mean(x), sd(x), skewness(x),
kurtosis(x), sd3(x), skew4(x), kurt5(x), dumbk(x))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## sum(x) 76 76 78 76 77 105 100 a
## mean(x) 153 154 158 155 156 249 100 a
## sd(x) 312 315 565 322 339 23243 100 ab
## skewness(x) 6867 6892 6949 6900 6946 7532 100 c
## kurtosis(x) 6763 6798 6991 6805 6854 19361 100 c
## sd3(x) 684 686 694 689 696 736 100 b
## skew4(x) 6701 6715 6809 6725 6779 9134 100 c
## kurt5(x) 12305 12333 12437 12349 12456 13447 100 d
## dumbk(x) 14147 14195 14596 14213 14283 31997 100 e
resdf <- checkit
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1747239 93 3.5e+06 186 2.3e+06 124
## Vcells 3588564 27 2.2e+07 168 3.4e+07 262
# weights
slow_sd <- function(x, w) {
n0 <- length(x)
mu <- weighted.mean(x, w = w)
sg <- sqrt(sum(w * (x - mu)^2)/(n0 - 1))
c(sg, mu, n0)
}
checkit <- microbenchmark(cent_moments(x, max_order = 4,
wts = w, na_rm = TRUE, normalize_wts = FALSE),
sd3(x, wts = w), slow_sd(x, w))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## cent_moments(x, max_order = 4, wts = w, na_rm = TRUE, normalize_wts = FALSE) 14061 14257 14978 14651 15537 18682 100 c
## sd3(x, wts = w) 1007 1036 1095 1058 1133 1395 100 a
## slow_sd(x, w) 943 1024 1726 1757 1893 6646 100 b
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1748655 93 3.5e+06 186 2.3e+06 124
## Vcells 3594655 28 1.4e+07 107 3.4e+07 262
set.seed(12345)
x1 <- runif(10000)
x2 <- runif(length(x1))
checkit <- microbenchmark(as.centsums(x1, 1), as.centsums(x1,
2), as.centsums(x1, 3), as.centsums(x1, 4))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## as.centsums(x1, 1) 74 75 77 77 79 84 100 a
## as.centsums(x1, 2) 111 113 118 116 119 274 100 b
## as.centsums(x1, 3) 729 735 754 749 764 914 100 c
## as.centsums(x1, 4) 1279 1322 1339 1330 1365 1433 100 d
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1749140 94 3.5e+06 186 2.3e+06 124
## Vcells 3618419 28 1.4e+07 107 3.4e+07 262
# join them together
max_ord <- 6L
obj1 <- as.centsums(x1, max_ord)
obj2 <- as.centsums(x2, max_ord)
obj3 <- as.centsums(c(x1, x2), max_ord)
checkit <- microbenchmark(c(obj1, obj2), obj3 %-% obj1)
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## c(obj1, obj2) 14 15 17 15 15 190 100 b
## obj3 %-% obj1 10 11 12 11 12 85 100 a
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1749819 94 3.5e+06 186 2.3e+06 124
## Vcells 3619708 28 1.4e+07 107 3.4e+07 262
max_ord <- 6L
rs1 <- cent_sums(x1, max_ord)
rs2 <- cent_sums(x2, max_ord)
rs3 <- cent_sums(c(x1, x2), max_ord)
checkit <- microbenchmark(join_cent_sums(rs1, rs2),
unjoin_cent_sums(rs3, rs2), unjoin_cent_sums(rs3,
rs1))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## join_cent_sums(rs1, rs2) 2.1 2.1 2.4 2.2 2.4 19.1 100 a
## unjoin_cent_sums(rs3, rs2) 2.0 2.1 2.2 2.2 2.3 2.9 100 a
## unjoin_cent_sums(rs3, rs1) 2.0 2.1 2.2 2.2 2.3 5.7 100 a
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1749898 94 3.5e+06 186 2.3e+06 124
## Vcells 3620471 28 1.4e+07 107 3.4e+07 262
set.seed(54321)
x1 <- matrix(rnorm(100 * 4), ncol = 4)
x2 <- matrix(rnorm(100 * 4), ncol = 4)
max_ord <- 2L
# join them together
mobj1 <- as.centcosums(x1, max_ord)
mobj2 <- as.centcosums(x2, max_ord)
mobj3 <- as.centcosums(rbind(x1, x2), max_ord)
alt3 <- c(mobj1, mobj2)
# unjoin them, with this one weird operator:
alt2 <- mobj3 %-% mobj1
alt1 <- mobj3 %-% mobj2
checkit <- microbenchmark(as.centcosums(x1, max_ord),
mobj3 %-% mobj1)
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## as.centcosums(x1, max_ord) 55 58 60 58 59 170 100 b
## mobj3 %-% mobj1 17 18 19 18 19 34 100 a
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1750017 94 3.5e+06 186 2.3e+06 124
## Vcells 3601115 28 1.4e+07 107 3.4e+07 262
set.seed(4422)
x <- rnorm(10000)
xpos <- runif(length(x)) + 1
xm <- matrix(x, ncol = 1)
xmps <- matrix(xpos, ncol = 1)
w <- runif(length(x))
dumb_zscore <- function(x, window) {
altz <- sapply(seq_along(x), function(iii) {
rowi <- max(1, iii - window + 1)
xrang <- x[rowi:iii]
(x[iii] - mean(xrang))/sd(xrang)
}, simplify = TRUE)
}
# run fun on each wins sized window...
silly_fun <- function(x, wins, fun, ...) {
xout <- rep(NA, length(x))
for (iii in seq_along(x)) {
xout[iii] <- fun(x[max(1, iii - wins + 1):iii],
...)
}
xout
}
wins <- 250
checkit <- microbenchmark(silly_fun(x, wins, sum, na.rm = FALSE),
silly_fun(x, wins, mean, na.rm = FALSE), running_sum(x,
wins), running_mean(x, wins), roll::roll_sum(xm,
wins), roll::roll_mean(xm, wins), roll::roll_sd(xm,
wins), RollingWindow::RollingSum(x, wins, na_method = "ignore"),
RollingWindow::RollingSum(x, wins), RollingWindow::RollingMean(x,
wins), RollingWindow::RollingStd(x, wins),
RcppRoll::roll_sum(xm, n = wins, align = "right",
fill = NA), RcppRoll::roll_mean(xm, n = wins,
align = "right", fill = NA), RcppRoll::roll_sd(xm,
n = wins, align = "right", fill = NA), running_sd(x,
wins, na_rm = FALSE, restart_period = 50000L),
running_sd(x, wins, na_rm = TRUE, restart_period = 1000L),
running_sd3(x, wins), running_skew(x, wins), running_skew4(x,
wins), running_kurt(x, wins), running_kurt5(x,
wins), running_tstat(x, wins), running_zscored(x,
wins), running_sharpe(x, wins), running_apx_median(x,
wins), running_centered(x, wins), running_scaled(x,
wins))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## silly_fun(x, wins, sum, na.rm = FALSE) 17074 18585 23216 20422 25782 118441 100 f
## silly_fun(x, wins, mean, na.rm = FALSE) 45077 48352 52609 51511 55202 78306 100 g
## running_sum(x, wins) 72 78 150 82 88 6792 100 a
## running_mean(x, wins) 69 76 82 78 86 123 100 a
## roll::roll_sum(xm, wins) 94 111 136 123 146 591 100 a
## roll::roll_mean(xm, wins) 102 117 139 135 156 203 100 a
## roll::roll_sd(xm, wins) 360 404 442 433 468 651 100 a
## RollingWindow::RollingSum(x, wins, na_method = "ignore") 251 280 381 301 328 7353 100 a
## RollingWindow::RollingSum(x, wins) 113 135 220 155 185 6095 100 a
## RollingWindow::RollingMean(x, wins) 146 167 251 183 217 5747 100 a
## RollingWindow::RollingStd(x, wins) 228 260 287 280 311 364 100 a
## RcppRoll::roll_sum(xm, n = wins, align = "right", fill = NA) 1645 1741 1864 1833 1944 2720 100 b
## RcppRoll::roll_mean(xm, n = wins, align = "right", fill = NA) 1649 1742 1854 1809 1945 2686 100 b
## RcppRoll::roll_sd(xm, n = wins, align = "right", fill = NA) 7964 8422 9973 8737 9353 17622 100 d
## running_sd(x, wins, na_rm = FALSE, restart_period = 50000L) 118 126 134 133 140 171 100 a
## running_sd(x, wins, na_rm = TRUE, restart_period = 1000L) 147 154 165 162 172 219 100 a
## running_sd3(x, wins) 302 320 398 332 347 6436 100 a
## running_skew(x, wins) 5068 5228 5516 5370 5743 7533 100 c
## running_skew4(x, wins) 5109 5359 5574 5522 5839 7166 100 c
## running_kurt(x, wins) 8217 8566 9025 8791 9352 15300 100 d
## running_kurt5(x, wins) 8758 9105 9555 9383 9952 13361 100 d
## running_tstat(x, wins) 357 378 397 392 413 470 100 a
## running_zscored(x, wins) 325 342 363 354 375 515 100 a
## running_sharpe(x, wins) 325 341 362 358 377 463 100 a
## running_apx_median(x, wins) 17657 18354 19319 19005 20045 25674 100 e
## running_centered(x, wins) 1095 1153 1210 1187 1278 1584 100 ab
## running_scaled(x, wins) 321 338 360 354 374 476 100 a
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1757929 94 3.5e+06 186 3.5e+06 186
## Vcells 3181449 24 1.4e+07 107 3.4e+07 262
library(readr)
FAKE_IT <- FALSE
if (FAKE_IT) {
resdf <- readr::read_csv("timings.csv")
# print(resdf)
} else {
readr::write_csv(resdf, "timings.csv")
}
performance regressions?
Can we see them here? load all the timing data to check.
library(magrittr)
library(dplyr)
library(readr)
library(tidyr)
library(knitr)
mysernum <- as.character(packageVersion("fromo"))
allt <- data.frame(fname = dir(".", "*.csv"), stringsAsFactors = FALSE) %>%
dplyr::filter(grepl("^timings_\\d.+\\d+.csv$",
fname)) %>% group_by(fname) %>% mutate(tims = list(readr::read_csv(fname))) %>%
ungroup() %>% tidyr::unnest() %>% mutate(sernum = gsub("^timings_(.+).csv$",
"\\1", fname)) %>% dplyr::select(-fname) %>% rbind(resdf %>%
dplyr::mutate(sernum = mysernum)) %>% group_by(sernum,
expr) %>% summarize(meantime = mean(time, na.rm = TRUE)) %>%
ungroup() %>% group_by(sernum) %>% mutate(sumx_time = median(ifelse(grepl("^sum\\(x\\)$",
expr), meantime, NA), na.rm = TRUE)) %>% ungroup() %>%
mutate(normalized = meantime/sumx_time) %>% arrange(sernum) %>%
group_by(expr) %>% mutate(first_norm = first(normalized),
last_norm = last(normalized)) %>% ungroup() %>%
mutate(relchange = normalized/first_norm, last_status = last_norm/first_norm) %>%
mutate(exgrp = ifelse(grepl("^(roll|RollingWindow|RcppRoll)::",
expr), "brand_x", ifelse(grepl("^running_",
expr), "running", "summarizing")))
library(ggplot2)
ph <- allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
ggplot(aes(sernum, normalized, group = expr, color = expr)) +
geom_line() + geom_point() + scale_y_log10() +
guides(colour = FALSE) + theme(axis.text.x = element_text(angle = -90,
hjust = 0)) + facet_grid(exgrp ~ ., scales = "free") +
labs(x = "release", y = "mean time taken, relative to sum(x)",
title = "fromo microbenchmark timings, lasagna")
print(ph)
ph <- allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
ggplot(aes(sernum, relchange, group = expr, color = expr)) +
geom_line() + geom_point() + scale_y_log10() +
guides(colour = FALSE) + theme(axis.text.x = element_text(angle = -90,
hjust = 0)) + facet_grid(exgrp ~ ., scales = "free") +
labs(x = "release", y = "normalized time taken, relative to first iteration",
title = "fromo microbenchmark timings, spaghetti")
print(ph)
ph <- allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
ggplot(aes(sernum, relchange)) + geom_boxplot(aes(group = sernum),
alpha = 0.7) + stat_summary(aes(group = "1", color = "mean"),
fun.y = mean, geom = "line") + scale_y_log10() +
theme(axis.text.x = element_text(angle = -90, hjust = 0)) +
facet_grid(exgrp ~ ., scales = "free") + labs(x = "release",
y = "normalized time taken, relative to first iteration",
color = "stat", title = "fromo microbenchmark timings, boxplots")
print(ph)
allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
select(expr, sernum, relchange, last_status) %>%
tidyr::spread(key = "sernum", value = "relchange") %>%
arrange(desc(last_status)) %>% select(-last_status) %>%
head(n = 50) %>% kable()
|expr | 0.1.3.3001| 0.1.3.3200| 0.1.3.3330| 0.1.3.3400| 0.1.3.4000| 0.1.3.4100| 0.1.3.4200| 0.1.3.4300| 0.1.3.4400| 0.1.3.4500| 0.1.3.4510| 0.1.3.5000| 0.1.3.5010| 0.1.3.5020| 0.1.3.6000| 0.1.3.6100| 0.1.3.6200| 0.1.3.6300| 0.1.3.6500| 0.1.3.6510| 0.1.3.6520| 0.1.3.6530| 0.1.3.6540| 0.1.3.6550| 0.1.3.6560| 0.1.3.6600| 0.1.3.6660| 0.1.3.6661| 0.1.3.6662| 0.1.3.6665| 0.1.3.6666| 0.1.3.6667| 0.1.3.6668| 0.1.3.6701| 0.1.3.6704| 0.1.3.6790| 0.1.3.6793| 0.1.3.6794| 0.1.3.6796|
|:----------------------------------------------------------------------------|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|
|running_sum(x, wins) | 1| NA| NA| NA| 2.72| 2.18| 1.95| NA| 2.02| 2.53| 2.74| 2.45| 3.23| 3.72| 2.50| 2.70| 2.74| 2.74| 2.79| 2.74| 2.76| 2.76| 2.84| 2.70| 2.77| 3.31| 2.76| 2.74| 3.29| 3.14| 2.80| 3.10| 2.85| 3.17| 2.78| 1.96| 2.10| 2.02| 3.83|
|running_centered(x, wins) | 1| 0.93| 0.91| 0.91| 1.01| 0.98| 0.92| NA| 0.94| 0.84| 0.91| 0.87| 1.08| 1.06| 0.94| 1.04| 0.95| 0.95| 1.56| 1.61| 1.94| 1.86| 2.03| 2.05| 2.02| 2.04| 1.96| 1.89| 1.82| 1.94| 2.03| 1.83| 1.80| 0.99| 1.45| 1.58| 1.74| 1.66| 2.03|
|running_skew(x, wins) | NA| NA| NA| NA| NA| 1.00| 1.09| NA| 1.16| 1.05| 1.18| 1.04| 1.17| 1.18| 1.05| 1.18| 1.13| 1.13| 1.29| 1.29| 1.36| 1.35| 1.39| 1.38| 1.34| 1.39| 1.93| 1.93| 1.72| 1.75| 1.70| 1.19| 1.17| 1.04| 1.43| 1.60| 1.68| 1.65| 1.79|
|running_kurt(x, wins) | NA| NA| NA| NA| NA| 1.00| 0.91| NA| 0.97| 0.85| 0.95| 0.86| 0.97| 0.96| 0.86| 0.96| 0.92| 0.92| 1.05| 1.02| 1.06| 1.08| 1.09| 1.08| 1.06| 1.09| 1.04| 1.03| 0.97| 0.95| 0.93| 0.95| 0.92| 0.86| 1.26| 1.43| 1.50| 1.43| 1.65|
|running_kurt5(x, wins) | 1| 0.94| 0.91| 0.91| 1.01| 1.03| 0.93| NA| 1.00| 0.90| 0.98| 0.89| 1.13| 1.13| 1.01| 1.13| 1.09| 1.09| 1.13| 1.11| 1.11| 1.12| 1.09| 1.09| 1.10| 1.12| 1.05| 1.05| 0.96| 0.94| 0.92| 0.97| 0.94| 0.88| 1.26| 1.41| 1.45| 1.45| 1.61|
|running_skew4(x, wins) | 1| 0.94| 0.89| 0.89| 1.02| 1.03| 0.92| NA| 0.97| 0.87| 0.96| 0.86| 1.17| 1.18| 1.04| 1.17| 1.13| 1.13| 1.15| 1.13| 1.16| 1.19| 1.19| 1.16| 1.13| 1.17| 1.62| 1.61| 1.41| 1.45| 1.40| 0.99| 0.98| 0.87| 1.18| 1.34| 1.40| 1.36| 1.49|
|running_apx_median(x, wins) | 1| 0.92| 0.90| 0.90| 1.06| 1.09| 0.93| NA| 1.00| 0.89| 1.00| 0.88| 1.05| 1.05| 0.93| 1.05| 1.02| 1.02| 1.04| 1.00| 1.02| 1.03| 1.01| 1.01| 1.01| 1.04| 1.01| 1.01| 1.01| 1.08| 0.99| 1.08| 1.06| 1.06| 1.14| 1.25| 1.31| 1.27| 1.39|
|sd(x) | 1| 1.02| 1.01| 1.01| 1.01| 1.01| 1.03| 0.99| 1.01| 1.02| 1.00| 1.02| 1.02| 1.02| 1.04| 1.02| 0.69| 0.69| 1.01| 0.99| 0.99| 1.00| 1.30| 1.49| 0.99| 1.04| 1.00| 1.03| 0.77| 0.75| 0.78| 0.73| 0.95| 0.75| 0.79| 0.74| 0.78| 0.74| 1.25|
|sd3(x, wts = w) | 1| 0.94| 0.94| 0.94| 0.96| 0.97| 0.90| 0.90| 0.95| 0.90| 0.97| 0.83| 1.63| 1.56| 1.50| 1.54| 1.07| 1.07| 1.07| 1.05| 1.04| 1.06| 1.03| 1.10| 1.05| 0.97| 0.97| 0.92| 0.96| 0.94| 0.86| 0.93| 0.78| 0.81| 0.71| 0.79| 0.81| 0.79| 1.08|
|as.centcosums(x1, max_ord) | 1| 0.93| 1.02| 1.02| 0.96| 1.24| 0.98| 1.02| 0.98| 0.87| 1.13| 0.85| 1.12| 1.16| 0.83| 0.93| 0.92| 0.92| 1.06| 0.91| 1.07| 0.93| 0.94| 0.96| 0.98| 0.95| 1.00| 0.86| 0.87| 0.91| 0.91| 0.92| 0.93| 0.88| 0.78| 0.91| 0.99| 0.92| 1.03|
|sum(x) | 1| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00|
|mean(x) | 1| 1.02| 1.00| 1.00| 1.01| 1.01| 1.01| 1.01| 1.01| 1.06| 0.97| 1.03| 1.02| 1.00| 1.05| 1.02| 1.00| 1.00| 1.00| 0.98| 0.98| 0.99| 1.02| 1.01| 0.98| 1.01| 0.97| 1.02| 1.00| 1.01| 1.03| 0.97| 0.99| 1.00| 1.01| 0.98| 1.02| 0.98| 0.99|
|unjoin_cent_sums(rs3, rs1) | 1| 1.21| 0.97| 0.97| 0.91| 1.34| 0.88| 0.99| 0.97| 0.99| 0.98| 0.85| 1.10| 1.19| 0.99| 1.11| 1.09| 1.09| 1.08| 1.04| 0.94| 1.05| 0.92| 1.20| 1.12| 1.00| 1.03| 0.90| 1.08| 0.99| 1.22| 1.00| 1.01| 1.06| 0.91| 0.96| 1.15| 1.00| 0.98|
|join_cent_sums(rs1, rs2) | 1| 1.12| 0.96| 0.96| 1.02| 1.25| 1.12| 0.97| 1.10| 1.03| 1.05| 0.93| 1.02| 1.14| 1.01| 1.11| 1.16| 1.16| 1.11| 1.03| 1.13| 1.04| 1.06| 1.25| 1.05| 1.08| 1.15| 1.05| 1.16| 0.97| 1.18| 1.02| 1.03| 1.09| 0.90| 0.98| 1.11| 0.99| 0.98|
|cent_moments(x, max_order = 4, wts = w, na_rm = TRUE, normalize_wts = FALSE) | 1| 0.93| 0.94| 0.94| 1.01| 1.07| 0.96| 0.95| 1.02| 0.96| 1.06| 0.91| 1.04| 1.01| 0.97| 1.00| 0.96| 0.96| 1.02| 1.00| 1.01| 1.03| 0.98| 1.04| 1.01| 1.03| 1.03| 1.00| 0.90| 0.86| 0.79| 0.86| 0.84| 0.87| 0.78| 0.87| 0.89| 0.86| 0.96|
|mobj3 %-% mobj1 | 1| 0.93| 1.03| 1.03| 0.97| 1.30| 0.95| 1.04| 0.98| 0.88| 1.17| 0.83| 1.12| 1.14| 0.88| 0.89| 0.93| 0.93| 1.05| 0.91| 1.05| 0.92| 0.94| 0.99| 1.03| 0.93| 0.98| 0.84| 0.82| 0.89| 0.86| 0.86| 0.88| 0.83| 0.77| 0.84| 0.90| 0.85| 0.94|
|kurt5(x) | 1| 1.02| 1.00| 1.00| 1.09| 1.10| 1.07| 1.06| 1.12| 1.09| 1.08| 1.08| 1.15| 1.16| 1.12| 1.14| 1.09| 1.09| 1.15| 1.12| 1.11| 1.12| 1.11| 1.14| 1.12| 1.14| 1.05| 1.04| 0.95| 0.92| 0.88| 0.98| 0.97| 0.99| 0.95| 0.98| 1.00| 0.96| 0.93|
|kurtosis(x) | 1| 1.00| 1.00| 1.00| 1.06| 1.06| 1.03| 1.00| 1.05| 1.04| 1.03| 1.03| 1.04| 1.07| 1.04| 1.04| 1.04| 1.04| 1.05| 1.03| 1.01| 1.03| 1.06| 1.09| 1.05| 1.06| 1.05| 1.02| 0.92| 0.91| 0.90| 0.92| 0.92| 0.93| 0.95| 0.92| 0.95| 0.92| 0.91|
|skew4(x) | 1| 1.01| 1.01| 1.01| 1.07| 1.07| 1.06| 1.05| 1.10| 1.08| 1.07| 1.09| 1.14| 1.14| 1.12| 1.13| 1.09| 1.09| 1.16| 1.14| 1.13| 1.13| 1.15| 1.16| 1.12| 1.15| 1.77| 1.76| 1.59| 1.55| 1.51| 1.01| 1.00| 1.01| 0.98| 1.00| 1.03| 0.99| 0.90|
|dumbk(x) | 1| 1.03| 1.01| 1.01| 1.04| 1.04| 1.01| 0.99| 1.04| 1.02| 1.02| 1.02| 1.04| 1.04| 1.03| 1.03| 0.96| 0.96| 1.04| 1.02| 1.02| 1.02| 1.02| 1.04| 1.02| 1.05| 1.03| 1.02| 0.89| 0.89| 0.89| 0.89| 0.89| 0.90| 0.89| 0.90| 0.92| 0.89| 0.90|
|skewness(x) | 1| 1.00| 1.00| 1.00| 1.06| 1.07| 1.04| 1.01| 1.06| 1.05| 1.03| 1.03| 1.06| 1.08| 1.06| 1.05| 0.98| 0.98| 1.08| 1.03| 1.03| 1.04| 1.04| 1.06| 1.03| 1.06| 1.04| 1.03| 0.90| 0.89| 0.88| 0.90| 0.89| 0.90| 0.88| 0.90| 0.92| 0.88| 0.89|
|c(obj1, obj2) | 1| 1.29| 0.97| 0.97| 0.93| 0.98| 1.04| 0.87| 0.95| 0.85| 0.85| 0.81| 0.89| 0.88| 0.83| 0.96| 0.93| 0.93| 0.88| 1.01| 0.87| 0.91| 0.92| 0.90| 0.91| 1.01| 1.03| 0.94| 0.87| 0.89| 1.37| 0.90| 0.90| 0.92| 0.80| 0.84| 0.92| 0.87| 0.87|
|as.centsums(x1, 4) | 1| 0.84| 0.86| 0.86| 0.92| 0.99| 0.91| 0.89| 0.93| 0.86| 0.90| 0.84| 0.95| 0.95| 0.96| 0.95| 0.92| 0.92| 0.96| 0.96| 0.97| 0.98| 1.00| 0.91| 0.97| 0.94| 0.93| 0.90| 0.82| 0.80| 0.70| 0.84| 0.84| 0.85| 0.75| 0.84| 0.83| 0.87| 0.85|
|unjoin_cent_sums(rs3, rs2) | 1| 0.86| 0.75| 0.75| 0.87| 1.05| 0.95| 0.77| 0.78| 0.82| 0.84| 0.73| 0.81| 0.99| 0.78| 0.92| 0.93| 0.93| 0.86| 0.86| 0.81| 0.87| 0.84| 1.03| 0.83| 0.88| 1.05| 0.82| 0.92| 0.84| 1.08| 0.87| 0.83| 0.91| 0.72| 0.83| 0.92| 0.85| 0.85|
|sd3(x) | 1| 1.03| 1.01| 1.01| 1.03| 1.03| 1.02| 1.00| 1.02| 1.01| 0.99| 1.00| 1.77| 1.80| 1.71| 1.76| 1.34| 1.34| 1.33| 1.30| 1.31| 1.31| 1.31| 1.35| 1.01| 1.01| 1.00| 0.99| 1.00| 0.98| 0.97| 1.00| 0.80| 0.83| 0.76| 0.82| 0.83| 0.81| 0.85|
|obj3 %-% obj1 | 1| 1.53| 0.96| 0.96| 0.91| 0.97| 1.00| 0.87| 1.00| 0.85| 0.85| 0.79| 0.92| 0.95| 0.90| 0.90| 0.96| 0.96| 0.97| 1.02| 0.92| 0.91| 0.83| 0.90| 0.91| 0.98| 1.04| 0.86| 0.87| 0.89| 1.41| 0.88| 0.92| 0.89| 0.78| 0.84| 0.91| 0.86| 0.85|
|as.centsums(x1, 2) | 1| 0.82| 0.81| 0.81| 0.81| 0.88| 0.80| 0.78| 0.81| 0.77| 0.78| 0.72| 1.38| 1.38| 1.41| 1.38| 0.98| 0.98| 0.99| 0.98| 0.99| 1.02| 0.99| 0.94| 0.83| 0.82| 0.83| 0.78| 0.79| 0.80| 0.71| 0.79| 0.80| 0.79| 0.72| 0.80| 0.78| 0.81| 0.83|
|as.centsums(x1, 3) | 1| 0.85| 0.86| 0.86| 0.91| 0.96| 0.89| 0.87| 0.92| 0.84| 0.89| 0.83| 0.95| 0.95| 0.96| 0.94| 0.91| 0.91| 0.99| 0.97| 0.99| 1.00| 1.00| 0.93| 0.97| 0.96| 1.54| 1.46| 1.35| 1.31| 1.17| 0.84| 0.83| 0.83| 0.75| 0.84| 0.82| 0.85| 0.82|
|silly_fun(x, wins, mean, na.rm = FALSE) | 1| 0.91| 0.89| 0.89| 0.94| 0.94| 0.88| NA| 0.92| 0.82| 0.90| 0.80| 0.92| 0.88| 0.79| 0.88| 0.76| 0.76| 0.88| 0.88| 0.87| 0.89| 0.88| 0.87| 0.89| 0.92| 0.86| 0.90| 0.66| 0.71| 0.59| 0.67| 0.64| 0.65| 0.59| 0.65| 0.66| 0.66| 0.72|
|running_sum(x, wins, robust = FALSE) | NA| 1.00| 0.66| 0.66| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA|
|silly_fun(x, wins, sum, na.rm = FALSE) | 1| 0.92| 0.90| 0.90| 0.90| 0.90| 0.82| NA| 0.87| 0.77| 0.87| 0.76| 0.86| 0.88| 0.76| 0.83| 0.67| 0.67| 0.86| 0.84| 0.84| 0.85| 0.87| 0.86| 0.84| 0.86| 0.82| 0.86| 0.49| 0.60| 0.49| 0.54| 0.53| 0.53| 0.46| 0.51| 0.56| 0.55| 0.60|
|running_sd3(x, wins) | NA| NA| NA| NA| 1.00| 1.04| 0.86| NA| 0.95| 0.78| 0.88| 0.80| 2.14| 2.03| 1.88| 2.05| 2.03| 2.03| 1.56| 1.46| 1.85| 1.60| 1.76| 1.69| 1.44| 1.59| 1.43| 1.35| 1.47| 1.47| 1.22| 1.33| 1.23| 0.53| 0.40| 0.46| 0.47| 0.46| 0.59|
|running_tstat(x, wins) | 1| 0.92| 0.94| 0.94| 0.99| 1.01| 0.82| NA| 0.86| 0.80| 0.90| 0.76| 1.23| 1.23| 1.10| 1.14| 1.18| 1.18| 1.13| 1.13| 1.43| 1.50| 1.48| 1.44| 1.33| 1.28| 1.28| 1.35| 1.19| 1.38| 1.16| 1.31| 1.21| 0.55| 0.44| 0.50| 0.52| 0.51| 0.55|
|running_scaled(x, wins) | 1| 0.94| 0.91| 0.91| 0.98| 1.03| 0.88| NA| 0.92| 0.79| 0.88| 0.79| 1.19| 1.19| 1.08| 1.19| 1.15| 1.15| 1.15| 1.15| 1.53| 1.47| 1.48| 1.53| 1.27| 1.40| 1.31| 1.31| 1.14| 1.25| 1.09| 1.20| 1.12| 0.72| 0.41| 0.46| 0.48| 0.47| 0.51|
|running_sharpe(x, wins) | 1| 0.99| 0.90| 0.90| 0.98| 1.02| 0.84| NA| 0.90| 0.79| 0.88| 0.77| 1.20| 1.19| 1.05| 1.11| 1.13| 1.13| 1.11| 1.08| 1.55| 1.42| 1.40| 1.44| 1.28| 1.31| 1.29| 1.34| 1.12| 1.29| 1.11| 1.26| 1.19| 0.52| 0.40| 0.45| 0.47| 0.46| 0.51|
|running_zscored(x, wins) | 1| 0.90| 0.88| 0.88| 1.01| 0.98| 0.80| NA| 0.89| 0.77| 0.84| 0.78| 1.20| 1.17| 1.02| 1.14| 1.10| 1.10| 1.15| 1.09| 1.47| 1.41| 1.44| 1.42| 1.28| 1.29| 1.25| 1.26| 1.09| 1.29| 1.07| 1.20| 1.12| 0.51| 0.39| 0.44| 0.47| 0.44| 0.49|
|running_sum(x, wins, robust = TRUE) | NA| 1.00| 0.48| 0.48| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA|
|slow_sd(x, w) | 1| 0.94| 0.89| 0.89| 0.47| 0.49| 0.48| 0.47| 0.48| 0.47| 0.55| 0.44| 0.48| 0.47| 0.49| 0.46| 0.54| 0.54| 0.48| 0.47| 0.46| 0.46| 0.45| 0.49| 0.46| 0.49| 0.47| 0.46| 0.42| 0.41| 0.39| 0.39| 0.40| 0.40| 0.37| 0.40| 0.43| 0.41| 0.45|
|running_mean(x, wins) | 1| 0.92| 0.59| 0.59| 0.46| 0.47| 0.53| NA| 0.56| 0.66| 0.72| 0.52| 0.59| 0.59| 0.53| 0.70| 0.59| 0.59| 0.81| 0.77| 0.60| 0.71| 0.63| 0.70| 0.59| 0.62| 0.71| 0.60| 0.60| 0.64| 0.54| 0.60| 0.59| 0.60| 0.53| 0.80| 1.14| 0.82| 0.45|
|as.centsums(x1, 1) | 1| 0.81| 0.82| 0.82| 0.85| 0.92| 0.85| 0.81| 0.87| 0.83| 0.85| 0.77| 0.87| 0.85| 0.93| 0.87| 0.35| 0.35| 0.36| 0.35| 0.36| 0.36| 0.36| 0.34| 0.35| 0.36| 0.35| 0.34| 0.34| 0.34| 0.31| 0.34| 0.34| 0.34| 0.30| 0.34| 0.33| 0.35| 0.35|
|running_sd(x, wins, na_rm = TRUE, restart_period = 1000L) | NA| NA| NA| NA| 1.00| 1.03| 0.87| NA| 0.87| 0.76| 0.87| 0.75| 1.17| 1.08| 1.07| 0.96| 0.89| 0.89| 1.20| 1.24| 1.69| 1.79| 1.71| 1.64| 1.51| 1.52| 1.48| 1.61| 1.24| 1.41| 1.20| 1.31| 1.29| 0.43| 0.29| 2.34| 2.49| 2.49| 0.34|
|running_sd(x, wins, na_rm = FALSE, restart_period = 50000L) | NA| NA| NA| NA| 1.00| 1.03| 0.86| NA| 0.90| 0.85| 0.92| 0.83| 1.07| 1.06| 1.07| 1.02| 1.00| 1.00| 1.30| 1.35| 1.94| 2.04| 1.90| 1.90| 1.77| 1.74| 1.73| 1.81| 1.50| 1.71| 1.43| 1.55| 1.52| 0.42| 0.28| 0.31| 0.33| 0.32| 0.33|
allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
select(-sernum, -relchange, -meantime, -sumx_time,
-normalized) %>% distinct(expr, .keep_all = TRUE) %>%
arrange(desc(last_status)) %>% head(n = 20) %>%
kable()
|expr | first_norm| last_norm| last_status|exgrp |
|:----------------------------------------------------------------------------|----------:|---------:|-----------:|:-----------|
|running_sum(x, wins) | 0.51| 1.94| 3.83|running |
|running_centered(x, wins) | 7.68| 15.58| 2.03|running |
|running_skew(x, wins) | 39.66| 71.05| 1.79|running |
|running_kurt(x, wins) | 70.61| 116.25| 1.65|running |
|running_kurt5(x, wins) | 76.25| 123.07| 1.61|running |
|running_skew4(x, wins) | 48.07| 71.80| 1.49|running |
|running_apx_median(x, wins) | 179.28| 248.85| 1.39|running |
|sd(x) | 5.82| 7.27| 1.25|summarizing |
|sd3(x, wts = w) | 13.06| 14.10| 1.08|summarizing |
|as.centcosums(x1, max_ord) | 0.75| 0.77| 1.03|summarizing |
|sum(x) | 1.00| 1.00| 1.00|summarizing |
|mean(x) | 2.07| 2.04| 0.99|summarizing |
|unjoin_cent_sums(rs3, rs1) | 0.03| 0.03| 0.98|summarizing |
|join_cent_sums(rs1, rs2) | 0.03| 0.03| 0.98|summarizing |
|cent_moments(x, max_order = 4, wts = w, na_rm = TRUE, normalize_wts = FALSE) | 200.65| 192.94| 0.96|summarizing |
|mobj3 %-% mobj1 | 0.26| 0.24| 0.94|summarizing |
|kurt5(x) | 173.12| 160.21| 0.93|summarizing |
|kurtosis(x) | 98.46| 90.06| 0.91|summarizing |
|skew4(x) | 97.66| 87.71| 0.90|summarizing |
|dumbk(x) | 209.85| 188.01| 0.90|summarizing |
shabbychef/fromo documentation built on April 11, 2021, 11:03 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
fromo timings
To prevent performance regressions, compare them here, including benchmarks against other packages:
library(fromo)
library(RollingWindow)
library(roll)
library(RcppRoll)
library(microbenchmark)
library(moments)
library(RcppParallel)
# keep this constant for comparison
setThreadOptions(numThreads = 2)
print(Sys.info())
## sysname release version
## "Linux" "4.4.0-137-generic" "#163-Ubuntu SMP Mon Sep 24 13:14:43 UTC 2018"
## nodename machine login
## "2d2df85a7b70" "x86_64" "unknown"
## user effective_user
## "docker" "docker"
print(sessionInfo())
## R version 3.5.2 (2018-12-20)
## Platform: x86_64-pc-linux-gnu (64-bit)
## Running under: Debian GNU/Linux buster/sid
##
## Matrix products: default
## BLAS: /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.8.0
## LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.8.0
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8 LC_MONETARY=en_US.UTF-8
## [6] LC_MESSAGES=en_US.UTF-8 LC_PAPER=en_US.UTF-8 LC_NAME=C LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## attached base packages:
## [1] utils methods base
##
## other attached packages:
## [1] RcppParallel_4.4.2 RcppRoll_0.3.0 roll_1.1.2 RollingWindow_0.2 microbenchmark_1.4-6 moments_0.14 dplyr_0.7.8
## [8] ggplot2_3.1.0 knitr_1.21 fromo_0.1.3.6796
##
## loaded via a namespace (and not attached):
## [1] Rcpp_1.0.0 bindr_0.1.1 magrittr_1.5 grDevices_3.5.2 tidyselect_0.2.5 munsell_0.5.0 colorspace_1.3-2 R6_2.3.0 rlang_0.3.0.1
## [10] stringr_1.3.1 plyr_1.8.4 tools_3.5.2 grid_3.5.2 gtable_0.2.0 xfun_0.4 withr_2.1.2 stats_3.5.2 lazyeval_0.2.1
## [19] assertthat_0.2.0 tibble_1.4.2 crayon_1.3.4 bindrcpp_0.2.2 formatR_1.5 purrr_0.2.5 graphics_3.5.2 glue_1.3.0 evaluate_0.12
## [28] stringi_1.2.4 compiler_3.5.2 pillar_1.3.1 scales_1.0.0 pkgconfig_2.0.2
set.seed(12345)
x <- rnorm(1e+05)
xpos <- runif(length(x)) + 1
xm <- matrix(x, ncol = 1)
xmps <- matrix(xpos, ncol = 1)
w <- runif(length(x))
dumbk <- function(x) {
c(kurtosis(x) - 3, skewness(x), sd(x), mean(x),
length(x))
}
checkit <- microbenchmark(sum(x), mean(x), sd(x), skewness(x),
kurtosis(x), sd3(x), skew4(x), kurt5(x), dumbk(x))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## sum(x) 76 76 78 76 77 105 100 a
## mean(x) 153 154 158 155 156 249 100 a
## sd(x) 312 315 565 322 339 23243 100 ab
## skewness(x) 6867 6892 6949 6900 6946 7532 100 c
## kurtosis(x) 6763 6798 6991 6805 6854 19361 100 c
## sd3(x) 684 686 694 689 696 736 100 b
## skew4(x) 6701 6715 6809 6725 6779 9134 100 c
## kurt5(x) 12305 12333 12437 12349 12456 13447 100 d
## dumbk(x) 14147 14195 14596 14213 14283 31997 100 e
resdf <- checkit
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1747239 93 3.5e+06 186 2.3e+06 124
## Vcells 3588564 27 2.2e+07 168 3.4e+07 262
# weights
slow_sd <- function(x, w) {
n0 <- length(x)
mu <- weighted.mean(x, w = w)
sg <- sqrt(sum(w * (x - mu)^2)/(n0 - 1))
c(sg, mu, n0)
}
checkit <- microbenchmark(cent_moments(x, max_order = 4,
wts = w, na_rm = TRUE, normalize_wts = FALSE),
sd3(x, wts = w), slow_sd(x, w))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## cent_moments(x, max_order = 4, wts = w, na_rm = TRUE, normalize_wts = FALSE) 14061 14257 14978 14651 15537 18682 100 c
## sd3(x, wts = w) 1007 1036 1095 1058 1133 1395 100 a
## slow_sd(x, w) 943 1024 1726 1757 1893 6646 100 b
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1748655 93 3.5e+06 186 2.3e+06 124
## Vcells 3594655 28 1.4e+07 107 3.4e+07 262
set.seed(12345)
x1 <- runif(10000)
x2 <- runif(length(x1))
checkit <- microbenchmark(as.centsums(x1, 1), as.centsums(x1,
2), as.centsums(x1, 3), as.centsums(x1, 4))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## as.centsums(x1, 1) 74 75 77 77 79 84 100 a
## as.centsums(x1, 2) 111 113 118 116 119 274 100 b
## as.centsums(x1, 3) 729 735 754 749 764 914 100 c
## as.centsums(x1, 4) 1279 1322 1339 1330 1365 1433 100 d
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1749140 94 3.5e+06 186 2.3e+06 124
## Vcells 3618419 28 1.4e+07 107 3.4e+07 262
# join them together
max_ord <- 6L
obj1 <- as.centsums(x1, max_ord)
obj2 <- as.centsums(x2, max_ord)
obj3 <- as.centsums(c(x1, x2), max_ord)
checkit <- microbenchmark(c(obj1, obj2), obj3 %-% obj1)
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## c(obj1, obj2) 14 15 17 15 15 190 100 b
## obj3 %-% obj1 10 11 12 11 12 85 100 a
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1749819 94 3.5e+06 186 2.3e+06 124
## Vcells 3619708 28 1.4e+07 107 3.4e+07 262
max_ord <- 6L
rs1 <- cent_sums(x1, max_ord)
rs2 <- cent_sums(x2, max_ord)
rs3 <- cent_sums(c(x1, x2), max_ord)
checkit <- microbenchmark(join_cent_sums(rs1, rs2),
unjoin_cent_sums(rs3, rs2), unjoin_cent_sums(rs3,
rs1))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## join_cent_sums(rs1, rs2) 2.1 2.1 2.4 2.2 2.4 19.1 100 a
## unjoin_cent_sums(rs3, rs2) 2.0 2.1 2.2 2.2 2.3 2.9 100 a
## unjoin_cent_sums(rs3, rs1) 2.0 2.1 2.2 2.2 2.3 5.7 100 a
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1749898 94 3.5e+06 186 2.3e+06 124
## Vcells 3620471 28 1.4e+07 107 3.4e+07 262
set.seed(54321)
x1 <- matrix(rnorm(100 * 4), ncol = 4)
x2 <- matrix(rnorm(100 * 4), ncol = 4)
max_ord <- 2L
# join them together
mobj1 <- as.centcosums(x1, max_ord)
mobj2 <- as.centcosums(x2, max_ord)
mobj3 <- as.centcosums(rbind(x1, x2), max_ord)
alt3 <- c(mobj1, mobj2)
# unjoin them, with this one weird operator:
alt2 <- mobj3 %-% mobj1
alt1 <- mobj3 %-% mobj2
checkit <- microbenchmark(as.centcosums(x1, max_ord),
mobj3 %-% mobj1)
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## as.centcosums(x1, max_ord) 55 58 60 58 59 170 100 b
## mobj3 %-% mobj1 17 18 19 18 19 34 100 a
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1750017 94 3.5e+06 186 2.3e+06 124
## Vcells 3601115 28 1.4e+07 107 3.4e+07 262
set.seed(4422)
x <- rnorm(10000)
xpos <- runif(length(x)) + 1
xm <- matrix(x, ncol = 1)
xmps <- matrix(xpos, ncol = 1)
w <- runif(length(x))
dumb_zscore <- function(x, window) {
altz <- sapply(seq_along(x), function(iii) {
rowi <- max(1, iii - window + 1)
xrang <- x[rowi:iii]
(x[iii] - mean(xrang))/sd(xrang)
}, simplify = TRUE)
}
# run fun on each wins sized window...
silly_fun <- function(x, wins, fun, ...) {
xout <- rep(NA, length(x))
for (iii in seq_along(x)) {
xout[iii] <- fun(x[max(1, iii - wins + 1):iii],
...)
}
xout
}
wins <- 250
checkit <- microbenchmark(silly_fun(x, wins, sum, na.rm = FALSE),
silly_fun(x, wins, mean, na.rm = FALSE), running_sum(x,
wins), running_mean(x, wins), roll::roll_sum(xm,
wins), roll::roll_mean(xm, wins), roll::roll_sd(xm,
wins), RollingWindow::RollingSum(x, wins, na_method = "ignore"),
RollingWindow::RollingSum(x, wins), RollingWindow::RollingMean(x,
wins), RollingWindow::RollingStd(x, wins),
RcppRoll::roll_sum(xm, n = wins, align = "right",
fill = NA), RcppRoll::roll_mean(xm, n = wins,
align = "right", fill = NA), RcppRoll::roll_sd(xm,
n = wins, align = "right", fill = NA), running_sd(x,
wins, na_rm = FALSE, restart_period = 50000L),
running_sd(x, wins, na_rm = TRUE, restart_period = 1000L),
running_sd3(x, wins), running_skew(x, wins), running_skew4(x,
wins), running_kurt(x, wins), running_kurt5(x,
wins), running_tstat(x, wins), running_zscored(x,
wins), running_sharpe(x, wins), running_apx_median(x,
wins), running_centered(x, wins), running_scaled(x,
wins))
print(checkit)
## Unit: microseconds
## expr min lq mean median uq max neval cld
## silly_fun(x, wins, sum, na.rm = FALSE) 17074 18585 23216 20422 25782 118441 100 f
## silly_fun(x, wins, mean, na.rm = FALSE) 45077 48352 52609 51511 55202 78306 100 g
## running_sum(x, wins) 72 78 150 82 88 6792 100 a
## running_mean(x, wins) 69 76 82 78 86 123 100 a
## roll::roll_sum(xm, wins) 94 111 136 123 146 591 100 a
## roll::roll_mean(xm, wins) 102 117 139 135 156 203 100 a
## roll::roll_sd(xm, wins) 360 404 442 433 468 651 100 a
## RollingWindow::RollingSum(x, wins, na_method = "ignore") 251 280 381 301 328 7353 100 a
## RollingWindow::RollingSum(x, wins) 113 135 220 155 185 6095 100 a
## RollingWindow::RollingMean(x, wins) 146 167 251 183 217 5747 100 a
## RollingWindow::RollingStd(x, wins) 228 260 287 280 311 364 100 a
## RcppRoll::roll_sum(xm, n = wins, align = "right", fill = NA) 1645 1741 1864 1833 1944 2720 100 b
## RcppRoll::roll_mean(xm, n = wins, align = "right", fill = NA) 1649 1742 1854 1809 1945 2686 100 b
## RcppRoll::roll_sd(xm, n = wins, align = "right", fill = NA) 7964 8422 9973 8737 9353 17622 100 d
## running_sd(x, wins, na_rm = FALSE, restart_period = 50000L) 118 126 134 133 140 171 100 a
## running_sd(x, wins, na_rm = TRUE, restart_period = 1000L) 147 154 165 162 172 219 100 a
## running_sd3(x, wins) 302 320 398 332 347 6436 100 a
## running_skew(x, wins) 5068 5228 5516 5370 5743 7533 100 c
## running_skew4(x, wins) 5109 5359 5574 5522 5839 7166 100 c
## running_kurt(x, wins) 8217 8566 9025 8791 9352 15300 100 d
## running_kurt5(x, wins) 8758 9105 9555 9383 9952 13361 100 d
## running_tstat(x, wins) 357 378 397 392 413 470 100 a
## running_zscored(x, wins) 325 342 363 354 375 515 100 a
## running_sharpe(x, wins) 325 341 362 358 377 463 100 a
## running_apx_median(x, wins) 17657 18354 19319 19005 20045 25674 100 e
## running_centered(x, wins) 1095 1153 1210 1187 1278 1584 100 ab
## running_scaled(x, wins) 321 338 360 354 374 476 100 a
resdf <- rbind(resdf, checkit)
gc()
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1757929 94 3.5e+06 186 3.5e+06 186
## Vcells 3181449 24 1.4e+07 107 3.4e+07 262
library(readr)
FAKE_IT <- FALSE
if (FAKE_IT) {
resdf <- readr::read_csv("timings.csv")
# print(resdf)
} else {
readr::write_csv(resdf, "timings.csv")
}
performance regressions?
Can we see them here? load all the timing data to check.
library(magrittr)
library(dplyr)
library(readr)
library(tidyr)
library(knitr)
mysernum <- as.character(packageVersion("fromo"))
allt <- data.frame(fname = dir(".", "*.csv"), stringsAsFactors = FALSE) %>%
dplyr::filter(grepl("^timings_\\d.+\\d+.csv$",
fname)) %>% group_by(fname) %>% mutate(tims = list(readr::read_csv(fname))) %>%
ungroup() %>% tidyr::unnest() %>% mutate(sernum = gsub("^timings_(.+).csv$",
"\\1", fname)) %>% dplyr::select(-fname) %>% rbind(resdf %>%
dplyr::mutate(sernum = mysernum)) %>% group_by(sernum,
expr) %>% summarize(meantime = mean(time, na.rm = TRUE)) %>%
ungroup() %>% group_by(sernum) %>% mutate(sumx_time = median(ifelse(grepl("^sum\\(x\\)$",
expr), meantime, NA), na.rm = TRUE)) %>% ungroup() %>%
mutate(normalized = meantime/sumx_time) %>% arrange(sernum) %>%
group_by(expr) %>% mutate(first_norm = first(normalized),
last_norm = last(normalized)) %>% ungroup() %>%
mutate(relchange = normalized/first_norm, last_status = last_norm/first_norm) %>%
mutate(exgrp = ifelse(grepl("^(roll|RollingWindow|RcppRoll)::",
expr), "brand_x", ifelse(grepl("^running_",
expr), "running", "summarizing")))
library(ggplot2)
ph <- allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
ggplot(aes(sernum, normalized, group = expr, color = expr)) +
geom_line() + geom_point() + scale_y_log10() +
guides(colour = FALSE) + theme(axis.text.x = element_text(angle = -90,
hjust = 0)) + facet_grid(exgrp ~ ., scales = "free") +
labs(x = "release", y = "mean time taken, relative to sum(x)",
title = "fromo microbenchmark timings, lasagna")
print(ph)
ph <- allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
ggplot(aes(sernum, relchange, group = expr, color = expr)) +
geom_line() + geom_point() + scale_y_log10() +
guides(colour = FALSE) + theme(axis.text.x = element_text(angle = -90,
hjust = 0)) + facet_grid(exgrp ~ ., scales = "free") +
labs(x = "release", y = "normalized time taken, relative to first iteration",
title = "fromo microbenchmark timings, spaghetti")
print(ph)
ph <- allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
ggplot(aes(sernum, relchange)) + geom_boxplot(aes(group = sernum),
alpha = 0.7) + stat_summary(aes(group = "1", color = "mean"),
fun.y = mean, geom = "line") + scale_y_log10() +
theme(axis.text.x = element_text(angle = -90, hjust = 0)) +
facet_grid(exgrp ~ ., scales = "free") + labs(x = "release",
y = "normalized time taken, relative to first iteration",
color = "stat", title = "fromo microbenchmark timings, boxplots")
print(ph)
allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
select(expr, sernum, relchange, last_status) %>%
tidyr::spread(key = "sernum", value = "relchange") %>%
arrange(desc(last_status)) %>% select(-last_status) %>%
head(n = 50) %>% kable()
|expr | 0.1.3.3001| 0.1.3.3200| 0.1.3.3330| 0.1.3.3400| 0.1.3.4000| 0.1.3.4100| 0.1.3.4200| 0.1.3.4300| 0.1.3.4400| 0.1.3.4500| 0.1.3.4510| 0.1.3.5000| 0.1.3.5010| 0.1.3.5020| 0.1.3.6000| 0.1.3.6100| 0.1.3.6200| 0.1.3.6300| 0.1.3.6500| 0.1.3.6510| 0.1.3.6520| 0.1.3.6530| 0.1.3.6540| 0.1.3.6550| 0.1.3.6560| 0.1.3.6600| 0.1.3.6660| 0.1.3.6661| 0.1.3.6662| 0.1.3.6665| 0.1.3.6666| 0.1.3.6667| 0.1.3.6668| 0.1.3.6701| 0.1.3.6704| 0.1.3.6790| 0.1.3.6793| 0.1.3.6794| 0.1.3.6796| |:----------------------------------------------------------------------------|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:|----------:| |running_sum(x, wins) | 1| NA| NA| NA| 2.72| 2.18| 1.95| NA| 2.02| 2.53| 2.74| 2.45| 3.23| 3.72| 2.50| 2.70| 2.74| 2.74| 2.79| 2.74| 2.76| 2.76| 2.84| 2.70| 2.77| 3.31| 2.76| 2.74| 3.29| 3.14| 2.80| 3.10| 2.85| 3.17| 2.78| 1.96| 2.10| 2.02| 3.83| |running_centered(x, wins) | 1| 0.93| 0.91| 0.91| 1.01| 0.98| 0.92| NA| 0.94| 0.84| 0.91| 0.87| 1.08| 1.06| 0.94| 1.04| 0.95| 0.95| 1.56| 1.61| 1.94| 1.86| 2.03| 2.05| 2.02| 2.04| 1.96| 1.89| 1.82| 1.94| 2.03| 1.83| 1.80| 0.99| 1.45| 1.58| 1.74| 1.66| 2.03| |running_skew(x, wins) | NA| NA| NA| NA| NA| 1.00| 1.09| NA| 1.16| 1.05| 1.18| 1.04| 1.17| 1.18| 1.05| 1.18| 1.13| 1.13| 1.29| 1.29| 1.36| 1.35| 1.39| 1.38| 1.34| 1.39| 1.93| 1.93| 1.72| 1.75| 1.70| 1.19| 1.17| 1.04| 1.43| 1.60| 1.68| 1.65| 1.79| |running_kurt(x, wins) | NA| NA| NA| NA| NA| 1.00| 0.91| NA| 0.97| 0.85| 0.95| 0.86| 0.97| 0.96| 0.86| 0.96| 0.92| 0.92| 1.05| 1.02| 1.06| 1.08| 1.09| 1.08| 1.06| 1.09| 1.04| 1.03| 0.97| 0.95| 0.93| 0.95| 0.92| 0.86| 1.26| 1.43| 1.50| 1.43| 1.65| |running_kurt5(x, wins) | 1| 0.94| 0.91| 0.91| 1.01| 1.03| 0.93| NA| 1.00| 0.90| 0.98| 0.89| 1.13| 1.13| 1.01| 1.13| 1.09| 1.09| 1.13| 1.11| 1.11| 1.12| 1.09| 1.09| 1.10| 1.12| 1.05| 1.05| 0.96| 0.94| 0.92| 0.97| 0.94| 0.88| 1.26| 1.41| 1.45| 1.45| 1.61| |running_skew4(x, wins) | 1| 0.94| 0.89| 0.89| 1.02| 1.03| 0.92| NA| 0.97| 0.87| 0.96| 0.86| 1.17| 1.18| 1.04| 1.17| 1.13| 1.13| 1.15| 1.13| 1.16| 1.19| 1.19| 1.16| 1.13| 1.17| 1.62| 1.61| 1.41| 1.45| 1.40| 0.99| 0.98| 0.87| 1.18| 1.34| 1.40| 1.36| 1.49| |running_apx_median(x, wins) | 1| 0.92| 0.90| 0.90| 1.06| 1.09| 0.93| NA| 1.00| 0.89| 1.00| 0.88| 1.05| 1.05| 0.93| 1.05| 1.02| 1.02| 1.04| 1.00| 1.02| 1.03| 1.01| 1.01| 1.01| 1.04| 1.01| 1.01| 1.01| 1.08| 0.99| 1.08| 1.06| 1.06| 1.14| 1.25| 1.31| 1.27| 1.39| |sd(x) | 1| 1.02| 1.01| 1.01| 1.01| 1.01| 1.03| 0.99| 1.01| 1.02| 1.00| 1.02| 1.02| 1.02| 1.04| 1.02| 0.69| 0.69| 1.01| 0.99| 0.99| 1.00| 1.30| 1.49| 0.99| 1.04| 1.00| 1.03| 0.77| 0.75| 0.78| 0.73| 0.95| 0.75| 0.79| 0.74| 0.78| 0.74| 1.25| |sd3(x, wts = w) | 1| 0.94| 0.94| 0.94| 0.96| 0.97| 0.90| 0.90| 0.95| 0.90| 0.97| 0.83| 1.63| 1.56| 1.50| 1.54| 1.07| 1.07| 1.07| 1.05| 1.04| 1.06| 1.03| 1.10| 1.05| 0.97| 0.97| 0.92| 0.96| 0.94| 0.86| 0.93| 0.78| 0.81| 0.71| 0.79| 0.81| 0.79| 1.08| |as.centcosums(x1, max_ord) | 1| 0.93| 1.02| 1.02| 0.96| 1.24| 0.98| 1.02| 0.98| 0.87| 1.13| 0.85| 1.12| 1.16| 0.83| 0.93| 0.92| 0.92| 1.06| 0.91| 1.07| 0.93| 0.94| 0.96| 0.98| 0.95| 1.00| 0.86| 0.87| 0.91| 0.91| 0.92| 0.93| 0.88| 0.78| 0.91| 0.99| 0.92| 1.03| |sum(x) | 1| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| 1.00| |mean(x) | 1| 1.02| 1.00| 1.00| 1.01| 1.01| 1.01| 1.01| 1.01| 1.06| 0.97| 1.03| 1.02| 1.00| 1.05| 1.02| 1.00| 1.00| 1.00| 0.98| 0.98| 0.99| 1.02| 1.01| 0.98| 1.01| 0.97| 1.02| 1.00| 1.01| 1.03| 0.97| 0.99| 1.00| 1.01| 0.98| 1.02| 0.98| 0.99| |unjoin_cent_sums(rs3, rs1) | 1| 1.21| 0.97| 0.97| 0.91| 1.34| 0.88| 0.99| 0.97| 0.99| 0.98| 0.85| 1.10| 1.19| 0.99| 1.11| 1.09| 1.09| 1.08| 1.04| 0.94| 1.05| 0.92| 1.20| 1.12| 1.00| 1.03| 0.90| 1.08| 0.99| 1.22| 1.00| 1.01| 1.06| 0.91| 0.96| 1.15| 1.00| 0.98| |join_cent_sums(rs1, rs2) | 1| 1.12| 0.96| 0.96| 1.02| 1.25| 1.12| 0.97| 1.10| 1.03| 1.05| 0.93| 1.02| 1.14| 1.01| 1.11| 1.16| 1.16| 1.11| 1.03| 1.13| 1.04| 1.06| 1.25| 1.05| 1.08| 1.15| 1.05| 1.16| 0.97| 1.18| 1.02| 1.03| 1.09| 0.90| 0.98| 1.11| 0.99| 0.98| |cent_moments(x, max_order = 4, wts = w, na_rm = TRUE, normalize_wts = FALSE) | 1| 0.93| 0.94| 0.94| 1.01| 1.07| 0.96| 0.95| 1.02| 0.96| 1.06| 0.91| 1.04| 1.01| 0.97| 1.00| 0.96| 0.96| 1.02| 1.00| 1.01| 1.03| 0.98| 1.04| 1.01| 1.03| 1.03| 1.00| 0.90| 0.86| 0.79| 0.86| 0.84| 0.87| 0.78| 0.87| 0.89| 0.86| 0.96| |mobj3 %-% mobj1 | 1| 0.93| 1.03| 1.03| 0.97| 1.30| 0.95| 1.04| 0.98| 0.88| 1.17| 0.83| 1.12| 1.14| 0.88| 0.89| 0.93| 0.93| 1.05| 0.91| 1.05| 0.92| 0.94| 0.99| 1.03| 0.93| 0.98| 0.84| 0.82| 0.89| 0.86| 0.86| 0.88| 0.83| 0.77| 0.84| 0.90| 0.85| 0.94| |kurt5(x) | 1| 1.02| 1.00| 1.00| 1.09| 1.10| 1.07| 1.06| 1.12| 1.09| 1.08| 1.08| 1.15| 1.16| 1.12| 1.14| 1.09| 1.09| 1.15| 1.12| 1.11| 1.12| 1.11| 1.14| 1.12| 1.14| 1.05| 1.04| 0.95| 0.92| 0.88| 0.98| 0.97| 0.99| 0.95| 0.98| 1.00| 0.96| 0.93| |kurtosis(x) | 1| 1.00| 1.00| 1.00| 1.06| 1.06| 1.03| 1.00| 1.05| 1.04| 1.03| 1.03| 1.04| 1.07| 1.04| 1.04| 1.04| 1.04| 1.05| 1.03| 1.01| 1.03| 1.06| 1.09| 1.05| 1.06| 1.05| 1.02| 0.92| 0.91| 0.90| 0.92| 0.92| 0.93| 0.95| 0.92| 0.95| 0.92| 0.91| |skew4(x) | 1| 1.01| 1.01| 1.01| 1.07| 1.07| 1.06| 1.05| 1.10| 1.08| 1.07| 1.09| 1.14| 1.14| 1.12| 1.13| 1.09| 1.09| 1.16| 1.14| 1.13| 1.13| 1.15| 1.16| 1.12| 1.15| 1.77| 1.76| 1.59| 1.55| 1.51| 1.01| 1.00| 1.01| 0.98| 1.00| 1.03| 0.99| 0.90| |dumbk(x) | 1| 1.03| 1.01| 1.01| 1.04| 1.04| 1.01| 0.99| 1.04| 1.02| 1.02| 1.02| 1.04| 1.04| 1.03| 1.03| 0.96| 0.96| 1.04| 1.02| 1.02| 1.02| 1.02| 1.04| 1.02| 1.05| 1.03| 1.02| 0.89| 0.89| 0.89| 0.89| 0.89| 0.90| 0.89| 0.90| 0.92| 0.89| 0.90| |skewness(x) | 1| 1.00| 1.00| 1.00| 1.06| 1.07| 1.04| 1.01| 1.06| 1.05| 1.03| 1.03| 1.06| 1.08| 1.06| 1.05| 0.98| 0.98| 1.08| 1.03| 1.03| 1.04| 1.04| 1.06| 1.03| 1.06| 1.04| 1.03| 0.90| 0.89| 0.88| 0.90| 0.89| 0.90| 0.88| 0.90| 0.92| 0.88| 0.89| |c(obj1, obj2) | 1| 1.29| 0.97| 0.97| 0.93| 0.98| 1.04| 0.87| 0.95| 0.85| 0.85| 0.81| 0.89| 0.88| 0.83| 0.96| 0.93| 0.93| 0.88| 1.01| 0.87| 0.91| 0.92| 0.90| 0.91| 1.01| 1.03| 0.94| 0.87| 0.89| 1.37| 0.90| 0.90| 0.92| 0.80| 0.84| 0.92| 0.87| 0.87| |as.centsums(x1, 4) | 1| 0.84| 0.86| 0.86| 0.92| 0.99| 0.91| 0.89| 0.93| 0.86| 0.90| 0.84| 0.95| 0.95| 0.96| 0.95| 0.92| 0.92| 0.96| 0.96| 0.97| 0.98| 1.00| 0.91| 0.97| 0.94| 0.93| 0.90| 0.82| 0.80| 0.70| 0.84| 0.84| 0.85| 0.75| 0.84| 0.83| 0.87| 0.85| |unjoin_cent_sums(rs3, rs2) | 1| 0.86| 0.75| 0.75| 0.87| 1.05| 0.95| 0.77| 0.78| 0.82| 0.84| 0.73| 0.81| 0.99| 0.78| 0.92| 0.93| 0.93| 0.86| 0.86| 0.81| 0.87| 0.84| 1.03| 0.83| 0.88| 1.05| 0.82| 0.92| 0.84| 1.08| 0.87| 0.83| 0.91| 0.72| 0.83| 0.92| 0.85| 0.85| |sd3(x) | 1| 1.03| 1.01| 1.01| 1.03| 1.03| 1.02| 1.00| 1.02| 1.01| 0.99| 1.00| 1.77| 1.80| 1.71| 1.76| 1.34| 1.34| 1.33| 1.30| 1.31| 1.31| 1.31| 1.35| 1.01| 1.01| 1.00| 0.99| 1.00| 0.98| 0.97| 1.00| 0.80| 0.83| 0.76| 0.82| 0.83| 0.81| 0.85| |obj3 %-% obj1 | 1| 1.53| 0.96| 0.96| 0.91| 0.97| 1.00| 0.87| 1.00| 0.85| 0.85| 0.79| 0.92| 0.95| 0.90| 0.90| 0.96| 0.96| 0.97| 1.02| 0.92| 0.91| 0.83| 0.90| 0.91| 0.98| 1.04| 0.86| 0.87| 0.89| 1.41| 0.88| 0.92| 0.89| 0.78| 0.84| 0.91| 0.86| 0.85| |as.centsums(x1, 2) | 1| 0.82| 0.81| 0.81| 0.81| 0.88| 0.80| 0.78| 0.81| 0.77| 0.78| 0.72| 1.38| 1.38| 1.41| 1.38| 0.98| 0.98| 0.99| 0.98| 0.99| 1.02| 0.99| 0.94| 0.83| 0.82| 0.83| 0.78| 0.79| 0.80| 0.71| 0.79| 0.80| 0.79| 0.72| 0.80| 0.78| 0.81| 0.83| |as.centsums(x1, 3) | 1| 0.85| 0.86| 0.86| 0.91| 0.96| 0.89| 0.87| 0.92| 0.84| 0.89| 0.83| 0.95| 0.95| 0.96| 0.94| 0.91| 0.91| 0.99| 0.97| 0.99| 1.00| 1.00| 0.93| 0.97| 0.96| 1.54| 1.46| 1.35| 1.31| 1.17| 0.84| 0.83| 0.83| 0.75| 0.84| 0.82| 0.85| 0.82| |silly_fun(x, wins, mean, na.rm = FALSE) | 1| 0.91| 0.89| 0.89| 0.94| 0.94| 0.88| NA| 0.92| 0.82| 0.90| 0.80| 0.92| 0.88| 0.79| 0.88| 0.76| 0.76| 0.88| 0.88| 0.87| 0.89| 0.88| 0.87| 0.89| 0.92| 0.86| 0.90| 0.66| 0.71| 0.59| 0.67| 0.64| 0.65| 0.59| 0.65| 0.66| 0.66| 0.72| |running_sum(x, wins, robust = FALSE) | NA| 1.00| 0.66| 0.66| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| |silly_fun(x, wins, sum, na.rm = FALSE) | 1| 0.92| 0.90| 0.90| 0.90| 0.90| 0.82| NA| 0.87| 0.77| 0.87| 0.76| 0.86| 0.88| 0.76| 0.83| 0.67| 0.67| 0.86| 0.84| 0.84| 0.85| 0.87| 0.86| 0.84| 0.86| 0.82| 0.86| 0.49| 0.60| 0.49| 0.54| 0.53| 0.53| 0.46| 0.51| 0.56| 0.55| 0.60| |running_sd3(x, wins) | NA| NA| NA| NA| 1.00| 1.04| 0.86| NA| 0.95| 0.78| 0.88| 0.80| 2.14| 2.03| 1.88| 2.05| 2.03| 2.03| 1.56| 1.46| 1.85| 1.60| 1.76| 1.69| 1.44| 1.59| 1.43| 1.35| 1.47| 1.47| 1.22| 1.33| 1.23| 0.53| 0.40| 0.46| 0.47| 0.46| 0.59| |running_tstat(x, wins) | 1| 0.92| 0.94| 0.94| 0.99| 1.01| 0.82| NA| 0.86| 0.80| 0.90| 0.76| 1.23| 1.23| 1.10| 1.14| 1.18| 1.18| 1.13| 1.13| 1.43| 1.50| 1.48| 1.44| 1.33| 1.28| 1.28| 1.35| 1.19| 1.38| 1.16| 1.31| 1.21| 0.55| 0.44| 0.50| 0.52| 0.51| 0.55| |running_scaled(x, wins) | 1| 0.94| 0.91| 0.91| 0.98| 1.03| 0.88| NA| 0.92| 0.79| 0.88| 0.79| 1.19| 1.19| 1.08| 1.19| 1.15| 1.15| 1.15| 1.15| 1.53| 1.47| 1.48| 1.53| 1.27| 1.40| 1.31| 1.31| 1.14| 1.25| 1.09| 1.20| 1.12| 0.72| 0.41| 0.46| 0.48| 0.47| 0.51| |running_sharpe(x, wins) | 1| 0.99| 0.90| 0.90| 0.98| 1.02| 0.84| NA| 0.90| 0.79| 0.88| 0.77| 1.20| 1.19| 1.05| 1.11| 1.13| 1.13| 1.11| 1.08| 1.55| 1.42| 1.40| 1.44| 1.28| 1.31| 1.29| 1.34| 1.12| 1.29| 1.11| 1.26| 1.19| 0.52| 0.40| 0.45| 0.47| 0.46| 0.51| |running_zscored(x, wins) | 1| 0.90| 0.88| 0.88| 1.01| 0.98| 0.80| NA| 0.89| 0.77| 0.84| 0.78| 1.20| 1.17| 1.02| 1.14| 1.10| 1.10| 1.15| 1.09| 1.47| 1.41| 1.44| 1.42| 1.28| 1.29| 1.25| 1.26| 1.09| 1.29| 1.07| 1.20| 1.12| 0.51| 0.39| 0.44| 0.47| 0.44| 0.49| |running_sum(x, wins, robust = TRUE) | NA| 1.00| 0.48| 0.48| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| NA| |slow_sd(x, w) | 1| 0.94| 0.89| 0.89| 0.47| 0.49| 0.48| 0.47| 0.48| 0.47| 0.55| 0.44| 0.48| 0.47| 0.49| 0.46| 0.54| 0.54| 0.48| 0.47| 0.46| 0.46| 0.45| 0.49| 0.46| 0.49| 0.47| 0.46| 0.42| 0.41| 0.39| 0.39| 0.40| 0.40| 0.37| 0.40| 0.43| 0.41| 0.45| |running_mean(x, wins) | 1| 0.92| 0.59| 0.59| 0.46| 0.47| 0.53| NA| 0.56| 0.66| 0.72| 0.52| 0.59| 0.59| 0.53| 0.70| 0.59| 0.59| 0.81| 0.77| 0.60| 0.71| 0.63| 0.70| 0.59| 0.62| 0.71| 0.60| 0.60| 0.64| 0.54| 0.60| 0.59| 0.60| 0.53| 0.80| 1.14| 0.82| 0.45| |as.centsums(x1, 1) | 1| 0.81| 0.82| 0.82| 0.85| 0.92| 0.85| 0.81| 0.87| 0.83| 0.85| 0.77| 0.87| 0.85| 0.93| 0.87| 0.35| 0.35| 0.36| 0.35| 0.36| 0.36| 0.36| 0.34| 0.35| 0.36| 0.35| 0.34| 0.34| 0.34| 0.31| 0.34| 0.34| 0.34| 0.30| 0.34| 0.33| 0.35| 0.35| |running_sd(x, wins, na_rm = TRUE, restart_period = 1000L) | NA| NA| NA| NA| 1.00| 1.03| 0.87| NA| 0.87| 0.76| 0.87| 0.75| 1.17| 1.08| 1.07| 0.96| 0.89| 0.89| 1.20| 1.24| 1.69| 1.79| 1.71| 1.64| 1.51| 1.52| 1.48| 1.61| 1.24| 1.41| 1.20| 1.31| 1.29| 0.43| 0.29| 2.34| 2.49| 2.49| 0.34| |running_sd(x, wins, na_rm = FALSE, restart_period = 50000L) | NA| NA| NA| NA| 1.00| 1.03| 0.86| NA| 0.90| 0.85| 0.92| 0.83| 1.07| 1.06| 1.07| 1.02| 1.00| 1.00| 1.30| 1.35| 1.94| 2.04| 1.90| 1.90| 1.77| 1.74| 1.73| 1.81| 1.50| 1.71| 1.43| 1.55| 1.52| 0.42| 0.28| 0.31| 0.33| 0.32| 0.33|
allt %>% dplyr::filter(!grepl("brand_x", exgrp)) %>%
select(-sernum, -relchange, -meantime, -sumx_time,
-normalized) %>% distinct(expr, .keep_all = TRUE) %>%
arrange(desc(last_status)) %>% head(n = 20) %>%
kable()
|expr | first_norm| last_norm| last_status|exgrp | |:----------------------------------------------------------------------------|----------:|---------:|-----------:|:-----------| |running_sum(x, wins) | 0.51| 1.94| 3.83|running | |running_centered(x, wins) | 7.68| 15.58| 2.03|running | |running_skew(x, wins) | 39.66| 71.05| 1.79|running | |running_kurt(x, wins) | 70.61| 116.25| 1.65|running | |running_kurt5(x, wins) | 76.25| 123.07| 1.61|running | |running_skew4(x, wins) | 48.07| 71.80| 1.49|running | |running_apx_median(x, wins) | 179.28| 248.85| 1.39|running | |sd(x) | 5.82| 7.27| 1.25|summarizing | |sd3(x, wts = w) | 13.06| 14.10| 1.08|summarizing | |as.centcosums(x1, max_ord) | 0.75| 0.77| 1.03|summarizing | |sum(x) | 1.00| 1.00| 1.00|summarizing | |mean(x) | 2.07| 2.04| 0.99|summarizing | |unjoin_cent_sums(rs3, rs1) | 0.03| 0.03| 0.98|summarizing | |join_cent_sums(rs1, rs2) | 0.03| 0.03| 0.98|summarizing | |cent_moments(x, max_order = 4, wts = w, na_rm = TRUE, normalize_wts = FALSE) | 200.65| 192.94| 0.96|summarizing | |mobj3 %-% mobj1 | 0.26| 0.24| 0.94|summarizing | |kurt5(x) | 173.12| 160.21| 0.93|summarizing | |kurtosis(x) | 98.46| 90.06| 0.91|summarizing | |skew4(x) | 97.66| 87.71| 0.90|summarizing | |dumbk(x) | 209.85| 188.01| 0.90|summarizing |
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.