joshbode

fix_numbers = function(x) {

  # strip out irrelevant characters
  x = gsub('[^a-zA-Z0-9. ]', '', x)

  # map SI units
  si_map = c('k'=10^3, 'M'=10^6, 'G'=10^9)

  for (unit in names(si_map)) {
    zeros = paste0(rep('0', log10(si_map[unit])), collapse='')

joshbode

fix_numbers = function(x) {

  # strip out irrelevant characters
  x = gsub('[^a-zA-Z0-9. ]', '', x)

  # map SI units
  si_map = c('k'=10^3, 'M'=10^6, 'G'=10^9)

  for (unit in names(si_map)) {
    zeros = paste0(rep('0', log10(si_map[unit])), collapse='')

joshbode

# poor-man's rollup
# rollups like in SQL: GROUP BY ROLLUP(...)

# example:
#   d = data.frame(x=trunc(runif(25, 1, 10)), y=trunc(runif(25, 1, 5)), z=rnorm(25))
#   result = pmr(d, .(x, y, j=x / 2), summarise, p=sum(z), .labels=list(x='-900', y='Total'))

require(plyr)

pmr = function(.data, .variables, ...,

joshbode

#!/bin/sh

# gvim wrapper for git on cygwin and msysgit
# Josh Bode <[email protected]>
#
# .gitconfig setup:
# 
#     [core]
#         editor = gvim-git.sh
#

joshbode

#!/bin/bash

# wrapper for knitr

RMD_FILE="$1"
MD_FILE="/tmp/$(basename $0).$$.md"
HTML_FILE="${RMD_FILE%.*}.html"

read -d '' COMMAND << "EOF"
  library(knitr, quietly=TRUE)

joshbode

@echo off

setlocal enabledelayedexpansion

net use G: "\\host\share" /persistent:yes

sc stop themes
sc stop mssql$sqlexpress

:: register fonts as local user

joshbode

# coalesce missing values
# e.g.
# > x = c(NA, 2, NA, 4)
# > y = c(1, NA, NA, 4)
# > z = c(1, NA, 3, 4)
# > coalesce(x, y, z)
coalesce = function(x, ...) {
  result = x
  for (y in list(...)) {
    mask = is.na(result)

joshbode

# Path functions ported to R from Python os.path module

path_join = function(a, ...) {
  p = list(...)

  path = a
  for (b in p) {
    b_wins = 0  # set to 1 iff b makes path irrelevant
    if (path == "") {
      b_wins = 1

joshbode

#!/usr/bin/env python

# simple script to practice mental arithmetic

import random
import operator
import argparse
import re

answer_re = re.compile(r'^[-+]?\d+$')

joshbode

library(plyr)
library(ggplot2)

d = data.frame(YEAR=as.ordered(rep(c(2001:2010), 10)), val=rnorm(1000))

p = ggplot(data=join(d, ddply(d, .(YEAR), function(x) { quantile(x$val) }))) +
    geom_boxplot(aes(YEAR, val)) +
    geom_text(aes(YEAR, `25%`, label=round(`25%`, 2)), stat='boxplot', vjust=-0.8, size=2) +
    geom_text(aes(YEAR, `50%`, label=round(`50%`, 2)), stat='boxplot', vjust=-0.5, size=2) +
    geom_text(aes(YEAR, `75%`, label=round(`75%`, 2)), stat='boxplot', vjust=+1.0, size=2) +