00pen

It is written after reading Efficient R programming book website.(https://csgillespie.github.io/efficientR/)

1. Prerequisites

library("microbenchmark")
library("profvis")
library("ggplot2")

#install 3 packages and attach by using install.packages() and library()

2. profiling

profvis( expr = { any code })

# this code can show you how long each code takes and how much each code takes by percent

3. system and Ram

Sys.info()

#see your OS and its version etc.

# Note: uses 2+ GB RAM and several seconds or more depending on hardware

# 1: Create large dataset

X = as.data.frame(matrix(rnorm(1e8), nrow = 1e7))

# 2: Find the median of each column using a single core

r1 = lapply(X, median)

# 3: Find the median of each column using many cores

r2 = parallel::mclapply(X, median)

#mclapply only work in parallel on Mac or Linux. we must use parLapply() on Windows.

4. packages

install.packages('installr')

installr::updateR()

#if R return FALSE, your Rversion is uptodate.

pkgs = c('raster', 'leaflet', 'rgeos')

install.packages(pkgs)

inst = lapply(pkgs, library, characher.only = TRUE)

#to shorten code, install packages in only two line of code and use lapply

#to make library(pkgs[i]).

#we use library instead of require because the former returns error when pkg is not installed.

update.packages()

#default of ask parameter is TRUE, change it to FASLE if you want.

#if you want this code run automatically when starting R

#you can add update.packages(ask = FALSE) in .Rprofile startup file

1. Title 

... + ggtitle('title') # or labs(title = 'title', ...)

# title appears at topright by default. To move title to center

... + theme(plot.title = element_text(hjust = 0.5))

2. labels

... + xlabs('label') + ylabs('label') # or labs(x = 'xlab', y = 'ylab')

3. line

... + geom_line(size = 2, color = '') # just connect points with line

... + geom_abline(size = 2, color = '') # draw regression line

... + geom_vline(size = 2, color = '') # draw vertical line

... + geom_hline(size = 2, color = '') # draw horizontal line

4. draw dynamic plot with manipulate function

myplot <- function(dynamic variable){

d <- data.frame(..., temp = dynamic variable, ...),

g <- ggplot(d, aes(...))

g <- g + geom_line()

...

g

}

manipulate(myplot(dynamic variable), slider(min, max, step = 1))

1. coloring manual

1) geom_point, geom_line, etc.

grouping with color for discrete value

g <- ggplot(data = df)

g + geom_point(aes(x = var1,  y = var2, color = 'discrete variable name')) +

      geom_line(aes(x= var1, y = var2, color = 'discrete variable name'))+ 

      scale_color_manual('legend title', breaks = 'discrete variable name', values = 'color value')

# when drawing more than one point or line, breaks = c(), values = c()

grouping with color for continuous value

g <- ggplot(data = df)

g + geom_point(aes(x = var1,  y = var2, color = 'continuous variable name')) +

      geom_line(aes(x= var1, y = var2, color = 'continuous variable name'))+ 

      scale_color_continuous('legend title', low = 'color for low value', high = 'color for high value')

# scale_color_continuous can be exchanged with scale_color_gradient or scale_color_gradientn( n step gradient )

2) geom_bar, geom_hist, etc.

# almost same as above

... + scale_fill_manual('legend title', breaks = 'discrete variable name', values = 'color value')

# for coloring with discrete variable

... + scale_fill_continuous('legend title', low = 'color for low value', high = 'color for high value')

# for coloring with continuous variable