在 R 中创建多元 QQ 图

Question

I wish to test for multivariate and univariate normality with QQ plots in R. I have used several different methods - they all do not seem to work. I am new to R, and would be incredibly grateful for any help.

My data set = vdata_clean1

I have 15 predicting variables (IV's), but no response variable (DV) because I am running a Confirmatory Factor Analysis (the 15 variables represent 15 different types of items that make up the scale I am testing).

I have installed the following packages:

install.packages("lavaan", dependencies=TRUE)
install.packages ("semPlot")
install.packages ("semTools")
install.packages("psych")
install.packages("MVN")
install.packages("mvtnorm")
install.packages("ggplot2")
install.packages("qgraph")
install.packages("psych")

1. First, I used the mvn function (multivariate qq plot)

    mvn(data = vdata_clean1, subset = NULL, mvnTest = c("mardia"), covariance = TRUE, tol = 1e-25, alpha = 0.5, scale = FALSE, desc = TRUE, transform = "none", R = 1000, univariateTest = c("SW"), univariatePlot = "none", multivariatePlot = "qq", multivariateOutlierMethod = "none", bc = FALSE, bcType = "rounded", showOutliers = FALSE, showNewData = FALSE)

   => I receive no error messages but the plots do not show.

-

2. Second, I tried the qqnorm function (multivariate qq plot)

    qqnorm(vdata_clean1, ylim, main = "Normal QQ Plot", xlab = "Theoretical Quantiles", ylab = "Sample Quantiles", plot.it = TRUE, datax = FALSE)

   => I receive the following error message and no plots:

    Error: Must use a vector in `[`, not an object of class matrix. Call `rlang:last_error()` to see a backtrace'

-

3. I tried the ggplot function for one variable (ILproto_1) (univariate qq plot)

    ggplot(vdata_clean1, aes(sample = ILproto_1)) + stat_qq() + stat_qq_line() + labs(title= "Normality QQ Plot (ILproto_1)", y = "Sample Quantiles") + theme_classic() `

   => I received this error message:

    Don't know how to automatically pick scale for object of type haven_labelled. Defaulting to continuous. Don't know how to automatically pick scale for object of type haven_labelled. Defaulting to continuous. Error in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : polygon edge not found In addition: Warning message: In grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : no font could be found for family "Arial"

   -

4. I also tried a histogram with the ggplot function (for variable ILproto_1)

    ggp1 <- ggplot(vdata_clean1, aes(x = ILproto_1)) ggp1 <- ggp1 + geom_histogram(binwidth=1, colour="black", aes(y=..density.., fill=..count..)) ggp1 <- ggp1 + scale_fill_gradient("Count", low="#DCDCDC", high="#7C7C7C") ggp1 <- ggp1 + stat_function(fun=dnorm, color="red", args=list(mean=mean(vdata_clean1$ILproto_1), sd=sd(vdata_clean1$ILproto_1))) ggp1

=> There is no error message, but I receive no plot

Answer 1

It's difficult without a reproducible example. Still, this uses the mtcars dataset to create a dataframe with 11 variables with different variables using ggplot2 .

library(tidyr)
library(ggplot2)

mtcars%>%
  gather()%>%
  ggplot(., aes(sample = value)) +
  stat_qq()+
  facet_wrap(vars(key), scales ='free_y')

Answer 2

You can look into the mqqnorm function from the {RVAideMemoire} package.

It creates a multivariate QQplot based on squared generalized distances and uses chi-square as theoretical distribution.