How to generate correlation plot of my data.frame in R?

问题

It might be a simple question. I have a df and I want to generate a correlation plot for my data in R.

head(df)

            x  y
1 -0.10967469  1
2  1.06814661 93
3  0.71805993 46
4  0.60566332 84
5  0.73714006 12
6 -0.06029712  5

I've found a package called corPlot and I've generated two plots based on pearson & spearman methods.

corPlot(df, method = 'pearson')
corPlot(df, method = 'spearman')

here is my output with pearson method:

I wondered if there is another package to generate the same correlation plots that I am not aware of?

Thanks in advance,

回答1:

Try the corrplot library, here is an example from this link

library(corrplot)
M <- cor(mtcars)
corrplot(M, method = "circle")

回答2:

You should check out ?pairs. It's great for doing scatterplots for combinations of variables, but varying the type of graph that is displayed on the

lower triangle (pass function taking 2 variables to lower.panel argument)
- e.g., pairs(df, lower.panel=points for a scatter plot
diagonal (pass function taking 1 variable to diag.panel or text.panel argument)
- this is tricky! see help file for how to do a histogram
upper triangle (pass function taking 2 variables to upper.panel argument)
- e.g., pairs(df, upper.panel=function(x,y)text(0,0,cor(x,y))), but see below

Scatter plot and correlation, from help file (?pairs):

panel.cor <- function(x, y, digits = 2, prefix = "", cex.cor, ...)
{
    usr <- par("usr"); on.exit(par(usr))
    par(usr = c(0, 1, 0, 1))
    r <- abs(cor(x, y))
    txt <- format(c(r, 0.123456789), digits = digits)[1]
    txt <- paste0(prefix, txt)
    if(missing(cex.cor)) cex.cor <- 0.8/strwidth(txt)
    text(0.5, 0.5, txt, cex = cex.cor * r)
}
pairs(USJudgeRatings, lower.panel = panel.smooth, upper.panel = panel.cor)

With your data, do the following:

pairs(df, lower.panel=plot, upper.panel=panel.cor)

I love that function, and use it myself as-is. It might look a little odd with only an x and a y, but

回答3:

Try the corrlY Package

You can plot the correlation matrix using your data frame. Also, you can find a correlation using various methods.

see the example-

Methods of Correlation Using corrlY Package

Install Corrly

# install.packages("devtools")
devtools::install_github("maheshKulkarniX/corrlY")
library(plotly)
library(corrlY)

correlation Coefficient using Kendall Method

# Example: 
corr_coef_kendall(variable1= mtcars$cyl, variable2=mtcars$carb, decimal = 2) 
## [1] 0.47

correlation Coefficient using Pearson Method

 # Example: 
    corr_coef_pearson(variable1= mtcars$disp, variable2=mtcars$hp, decimal = 2)
    ## [1] 0.79

correlation Coefficient using spearman Method

# Example:
corr_coef_spearman(variable1= cars$speed, variable2=cars$dist, decimal = 2)
## [1] 0.83

correlation Matrix Plot Using corrlY

matrixly(data = mtcars)

Correlation Plot using corrlY

x <- c(0.10967469, 1.06814661, 0.71805993, 0.60566332, 0.73714006, -0.06029712)
y <- c(1, 93, 46, 84, 12, 5)
xy <- data.frame(x, y)
spearman<- corr_coef_spearman(variable1= xy$x, variable2=xy$y, decimal = 2)
corr_scatterly(data=xy,x= xy$x, y= xy$y,corr_coef=spearman, xname="x", yname="y")

For More Information visit the GitHub repository Data Visualization Package For All Types of Correlation Charts using Plotly Package.

For More examples visit corrlY Package Site corrlY package site

来源：https://stackoverflow.com/questions/31968022/how-to-generate-correlation-plot-of-my-data-frame-in-r

标签

plot

correlation

pearson