kevinushey

// [[Rcpp::depends(RcppGSL)]]

#include <gsl/gsl_randist.h>

#include <Rcpp.h>
using namespace Rcpp;

// [[Rcpp::export]]
NumericVector gsl_dexp(NumericVector x) {
  int n = x.size();

kevinushey

`|` <- function(x, y) {
  if (is.data.frame(x)) {
    return( eval(call("%.%", substitute(x), substitute(y)), envir=parent.frame()) )
  } else {
    return( base::"|"(x, y) )
  }
}

library(dplyr)
mtcars |

kevinushey

// run me by installing Rcpp and calling sourceCpp on this file

#include <Rcpp.h>
using namespace Rcpp;

// some initial stuff borrowed from R sources

#ifdef WORDS_BIGENDIAN
static const int hw = 0;
static const int lw = 1;

kevinushey

case2 <- function(n = 10, mu = 3, sigma = sqrt(5), p = 0.025, rep = 100){
  
  xbar <- vapply(1:rep, FUN.VALUE=numeric(1), function(i) {
    norm <- rnorm(mean = mu, sd = sigma, n = n)
    return(mean(norm))
  })
  
  low <- xbar - qnorm(1-p) * (sigma / sqrt(n))
  up <- xbar + qnorm(1-p) * (sigma / sqrt(n))
  rem <- mu > low & mu < up

kevinushey

git_status <- ''
class(git_status) <- "__git_status__"
print.__git_status__ <- function(x, ...) system("git status")

git_status ## prints 'git status' for the current dir

kevinushey

// This code copied and modified from the R sources (src/main/apply.c) for do_lapply.
// Essentially, we remove some of the op / arity checking, and then
// construct a call to FUN that also passes the index.

#define USE_RINTERNALS

#include <R.h>
#include <Rinternals.h>

// [[export]]

kevinushey

#include <Rcpp.h>
using namespace Rcpp;

union DoublePunner {
  double d;
  uint64_t p;
};

static DoublePunner NA_PUNNED = { NA_REAL };

kevinushey

#define debug

#ifdef DEBUG
#define debug(x) Rprintf(x)
#define debug2(x, y) Rprintf(x, y)
#define debug3(x, y, z) Rprintf(x, y, z)
#else
#define debug(x)
#define debug2(x, y)
#define debug3(x, y, z)

kevinushey

library(shiny)

shinyServer( function(input, output, session) {
  
})

kevinushey

Melting with Rcpp

A common data manipulation task is that of making 'wide' data 'long': for example, using the reshape2 library,

library(reshape2)
wide_df <- data.frame( stringsAsFactors=FALSE,