Charting Correlation Matrices in R

-
I noticed this very simple, very powerful article by James Marquez, Seven Easy Graphs to Visualize Correlation Matrices in R, in the Google+ community, R Programming for Data Analysis, so thought to give it a try, since I started some of my current analyses a decade ago by generating correlation matrices in Excel, which I've sometimes redone and improved in R.

Some of these packages are only designed for display, or as extensions to ggplot2:
These two are focused on more complex analysis:
As for data, I used Hofstede's culture dimensions, limited to developed countries. Using a broader and larger set of of countries would significantly reduce the correlations, in that only individuality and power distance would show a strong relationship, negative. The source data is here. 

 
 # Charting correlations in R
 # Source: http://jamesmarquezportfolio.com/correlation_matrices_in_r.html  
   
 # Clear workspace  
 rm(list = ls())  
   
 # Set working directory  
 setwd("../Data")  
 getwd()  

 # load data
 oecdData <- read.table("OECD - Quality of Life.csv", header = TRUE, sep = ",")  
 hofsted.vectors <- oecdData[,c('HofstederPowerDx', 'HofstederMasculinity', 'HofstederIndividuality', 'HofstederUncertaintyAvoidance', 'HofstederLongtermOrientation', 'HofstederIndulgence')]  

 #rename columns
 names(hofsted.vectors)[1:6] = c('PowerDx', 'Masculinity', 'Individuality', 'UAE', 'LTO', 'Indulgence')
  
 # PerformanceAnalytics  
 install.packages("PerformanceAnalytics", dependencies = TRUE)  
 library("PerformanceAnalytics")  
 chart.Correlation(hofsted.vectors, histogram = TRUE, pch = 19)  





# psych
 install.packages("psych", dependencies = TRUE)
 library(psych)
 pairs.panels(hofsted.vectors, scale = TRUE)





# corrplot  
 install.packages("corrplot", dependencies = TRUE)  
 library(corrplot)  
 corrplot.mixed(cor(hofsted.vectors), order = "hclust", tl.col = "black") 
 





# GGally  
 install.packages("GGally", dependencies = TRUE)  
 library(GGally)  
 ggpairs(hofsted.vectors)  





# ggcorrplot  
 install.packages("ggcorrplot", dependencies = TRUE)  
 library(ggcorrplot)  
 ggcorrplot(cor(hofsted.vectors), p.mat = cor_pmat(hofsted.vectors), hc.order = TRUE, type = 'lower')

Comments

Post a Comment

Popular posts from this blog

Developers in New York City by Zip Code

-
Image
Photo by Lukas Kloeppel from Pexels In 2016, after reading a Dice Insight article, I downloaded data that had technology professional numbers by zip code, along with density. A recent NY Times article How Big Tech Is Turning New York Into a Silicon Valley Rival prompted me to resurrect the data, decorate it a bit with neighborhood names, and then import it into Google Maps , which was surprisingly easy. The original data is available in Excel .
Read more

Cultural Dimensions and Coffee Consumption

-
Image
Photo by Viktoria Alipatova from Pexels Responding to a Treehugger article, Why Americans will never love tea as much as coffee , I initially wrote my personal preferences for tea and coffee , ending with, BTW, this has just given me an idea for comparing Hofstede's cultural dimensions and coffee and tea consumption. Afterward, I did some analysis in Excel, then ran the same processes in R using Visual Studio, then converted that to a Jupyter Notebook on Microsoft's Azure Notebooks . Although this analysis is limited to 45 countries that have Hofstede's Cultural Dimensions , as well as per capita consumption for both coffee and tea, it would seem that coffee consumption correlates with power distance, individuality, and masculinity. Tea had small correlations with the dimensions and sometimes in the same direction as coffee. A fuller analysis is available on Microsoft's Azure Notebook , but some quick findings: Higher power distance, lower coffee consumption
Read more