AnthonyEbert

# Courtesy of Miles McBain https://milesmcbain.xyz/hacking-r-library-paths

set_lib_paths <- function(lib_vec) {

  lib_vec <- normalizePath(lib_vec, mustWork = TRUE)

  shim_fun <- .libPaths
  shim_env <- new.env(parent = environment(shim_fun))
  shim_env$.Library <- character()
  shim_env$.Library.site <- character()

AnthonyEbert

Computational resources at Università della Svizzera italiana

CSCS

Webinar: https://youtu.be/dXCyCAgviZY?t=411

Email: help a-tea cscs.ch

AnthonyEbert

prob <- 0.1
len_z <- 1000
z <- rbinom(n = len_z, 1, prob)
n <- 1000

z_mat <- matrix(nrow = n, ncol = len_z)

for(i in 1:n){
  for(s in 1:len_z){
    oldz <- z[s]

AnthonyEbert

How to submit R code to a Portable Batch System (PBS) managed High Performance Computing (HPC) Facility (e.g. QUT's HPC Facility 'Lyra').

You will need:

1. An account with the relevant HPC facility (QUT staff and HDR students can request access to the QUT facility here),
2. the R Script (.R file) you want to run,
3. a Job Script (.sub file) that tells Lyra how to run your R script and,
4. a file containing any data your R Script needs to run.

Note: These instructions work on a system running Debian (GNU + Linux) with the Gnome Desktop Environment.

AnthonyEbert

# R hacks that are cool ------

## Function to create list with default elements

list_defaults <- function(
  a = 1,
  b = 3,
  c = "lsa"){

AnthonyEbert

Get old commit ready

Create folder in another repo directory e.g. repo-tester . Checkout the commit version that you want.

Put tex into single file

latexpand main.tex > output.tex

Do this for Git and repo-tester folders.

AnthonyEbert

Is subsampling y a problem?

The real data is y, the realisations (forward-simulations) are x. The distance d could refer to the MMD or the Wasserstein estimator.

Sampling strategies

1. Compute full d with y and x,
2. Draw a subsample from y and x and compute d,

AnthonyEbert

# For empirical posterior plots. Put the true values, repeated at the bottom two rows of the data. 

require(datasets)
data("swiss")
require(GGally)
require(ggplot2)

input_swiss <- rbind(swiss[,1:4], c(75, 60, 10, 15), c(75, 60, 10, 15))

my_low <- function(data, mapping, ...){

AnthonyEbert

library(queuecomputer)
 
arrivals <- cumsum(rexp(1e4))
service <- rexp(1e4, 0.6)
max_wait <- rexp(1e4, 0.01)
 
reneg <- TRUE

angry_output <- c()
 

AnthonyEbert

import rpy2
import rpy2.robjects as robjects
import numpy

# python3 to R ----------------------------

py_base_list_x = [1,6]
py_base_array_x = (1,6)
R_base_numeric_x = rpy2.robjects.vectors.FloatVector(py_base_list_x)