Must have R libraries

Original by reinie, 2016 

This summary note was Posted on

Here is a list of useful libraries (that will grow over time)

Making plots

library(Corrplot)  #to make nice correlations plots with values inside
library(ggplot2) #a must have for pretty graph, but can do without
library(reshape2) #sometimes needed to reshape data to plot with ggplot2

Data cleaning

library(Amelia) #to visualize missing data

Data input

library(readxl) #faster than read.xlsx for reading Excel files

Data manipulation

library(dplyr) #what made me use R, makes data manipulation more human
librar(plyr) #use it for merging data, simpler than dplyr on that one

Data statistics

library(moments) #to get ... moments
library(e1071) #to get skew and kurtosis and other stats from distribution
library(nortest) #to easily test normality

Programming

library(purr) #to use among other the map function to replace for loops

Performance for binary classification methods

library(InformationValue) #Get KS, AUROC, SomersD, confusionMatrix etc.. faster