R 进行主成分分析时出错
我正在做一个关于红酒质量的研究。数据集有12个变量,我的响应变量是R 进行主成分分析时出错,r,pca,R,Pca,我正在做一个关于红酒质量的研究。数据集有12个变量,我的响应变量是quality。我试图做一个主成分分析(PCA)。为此,我使用了如下所示的代码: wine <- read.csv("winequality-red.csv", sep = ";") q.pca <- prcomp(wine[, 1:11], center = TRUE, scale. = TRUE) plot(q.pca, type = "l", main = '') grid(nx = 10, ny = 14)
quality
。我试图做一个主成分分析(PCA)。为此,我使用了如下所示的代码:
wine <- read.csv("winequality-red.csv", sep = ";")
q.pca <- prcomp(wine[, 1:11], center = TRUE, scale. = TRUE)
plot(q.pca, type = "l", main = '')
grid(nx = 10, ny = 14)
title(main = "Principal component weight", sub = NULL, xlab = "components")
box()
在此之后,我得到了一个错误,其中提到:
错误:提供给离散刻度的连续值
请指出这段代码中的错误。Hmm,我没有收到提到的错误。相反,我在ggbiplot中得到了
错误(q.pca,choices=1:2,obs.scale=1,var.scale=1,:未找到对象“q.diag”
@Kim,尝试从上面运行完整的代码。q.pca是我用来存储上面的prcomp函数的变量。q.diag
对象缺失,如果注释掉'groups=q.diag
行,代码运行正常。您能提供对象吗?它是cle请注意,上面的代码不是生成图形的原因,因为图形的标题中有一个不在代码中的打字错误。
devtools::install_github("vqv/ggbiplot")
library("ggbiplot")
ggbiplot(
q.pca, choices = 1:2, obs.scale = 1, var.scale = 1,
groups = q.diag, ellipse = TRUE, circle = TRUE, varname.size = 4,
ellipse.prob = 0.68, circle.prob = 0.69
) +
scale_color_discrete(name = 'Quality (from 3 to 8)') +
theme_bw() +
labs(
title = "Principal Component Analysis",
subtitle = "1. Data distribution in the plan of PC1 and PC2\n2. Directions of components in the same plane"
) +
theme(legend.direction = 'horizontal', legend.position = 'bottom')