Laurae2

set.seed(1)
N <- 500000
p <- 100
pp <- 25
X <- matrix(runif(N * p), ncol = p)
betas <- 2 * runif(pp) - 1
sel <- sort(sample(p, pp))
m <- X[, sel] %*% betas - 1 + rnorm(N)
y <- rbinom(N, 1, plogis(m))

Laurae2

library(xgboost)

set.seed(1)
N <- 1000000
p <- 100
pp <- 25
X <- matrix(runif(N * p), ncol = p)
betas <- 2 * runif(pp) - 1
sel <- sort(sample(p, pp))
m <- X[, sel] %*% betas - 1 + rnorm(N)

Laurae2

library(xgboost)
library(lightgbm)
library(data.table)

setwd("/home/laurae/Documents/R/GBM-perf")

n_threads <- 16

data <- fread("HIGGS.csv")
labels <- data$V1

Laurae2

Walk on thin ice in R for gcc-7/g++-7 + CUDA 10.0.130

Note: OpenCL can run on Single Precision (gpu_use_dp = FALSE) or Double Precision (gpu_use_dp = TRUE) whereas CUDA is strictly Double Precision in LightGBM.

--

STEP 1: HACK IN SOME FILES

Hack in the following:
```

Laurae2

Code:

library(anytime)
library(flipTime)
library(data.table)

set.seed(1)
system.time({
  some_dates <- as.Date("2020-12-31") - sample(40000, 1e7, TRUE)