PCA_function function

Analysis: Principal components analysis

This function performs principal component analysis.

PCA_function(
  data,
  scale = TRUE,
  text = TRUE,
  pointsize = 5,
  textsize = 12,
  labelsize = 4,
  linesize = 0.6,
  repel = TRUE,
  ylab = NA,
  xlab = NA,
  groups = NA,
  sc = 1,
  font.family = "sans",
  theme = theme_bw(),
  label.legend = "Cluster",
  type.graph = "biplot"
)

Arguments

• data: Data.frame with data set. Line name must indicate the treatment
• scale: Performs data standardization (default is TRUE)
• text: Add label (default is TRUE)
• pointsize: Point size (default is 5)
• textsize: Text size (default is 12)
• labelsize: Label size (default is 4)
• linesize: Line size (default is 0.8)
• repel: Avoid text overlay (default is TRUE)
• ylab: Names y-axis
• xlab: Names x-axis
• groups: Define grouping
• sc: Secondary axis scale ratio (default is 1)
• font.family: Font family (default is sans)
• theme: Theme ggplot2 (default is theme_bw())
• label.legend: Legend title (when group is not NA)
• type.graph: Type of chart (default is biplot)

Returns

The eigenvalues and eigenvectors, the explanation percentages of each principal component, the correlations between the vectors with the principal components, as well as graphs are returned.

Details

The type.graph argument defines the graph that will be returned, in the case of "biplot" the biplot graph is returned with the first two main components and with eigenvalues and eigenvectors. In the case of "scores" only the treatment scores are returned, while for "cor" the correlations are returned. For "corPCA" a correlation between the vectors with the components is returned.

Examples

data(pomegranate)
medias=tabledesc(pomegranate)
PCA_function(medias)

Author(s)

Gabriel Danilo Shimizu