sjPlot (version 2.1.0)

sjp.pca: Plot PCA results

Description

Performes a principle component analysis on a data frame or matrix (with varimax rotation) and plots the factor solution as ellipses or tiles. In case a data frame is used as argument, the cronbach's alpha value for each factor scale will be calculated, i.e. all variables with the highest loading for a factor are taken for the reliability test. The result is an alpha value for each factor dimension.

Usage

sjp.pca(data, nmbr.fctr = NULL, fctr.load.tlrn = 0.1, plot.eigen = FALSE, digits = 2, title = NULL, axis.labels = NULL, type = c("bar", "circle", "tile"), geom.size = 0.6, geom.colors = "RdBu", wrap.title = 50, wrap.labels = 30, show.values = TRUE, show.cronb = TRUE, prnt.plot = TRUE)

Arguments

data

data.frame that should be used to compute a PCA, or a prcomp object.

nmbr.fctr

number of factors used for calculating the varimax rotation. By default, this value is NULL and the amount of factors is calculated according to the Kaiser-criteria.

fctr.load.tlrn

specifies the minimum difference a variable needs to have between factor loadings (components) in order to indicate a clear loading on just one factor and not diffusing over all factors. For instance, a variable with 0.8, 0.82 and 0.84 factor loading on 3 possible factors can not be clearly assigned to just one factor and thus would be removed from the principal component analysis. By default, the minimum difference of loading values between the highest and 2nd highest factor should be 0.1

plot.eigen

If TRUE, a plot showing the Eigenvalues according to the Kaiser criteria is plotted to determine the number of factors.

digits

numeric, amount of digits after decimal point when rounding estimates and values.

title

character vector, used as plot title. Depending on plot type and function, will be set automatically. If title = "", no title is printed.

axis.labels

character vector with labels used as axis labels. Optional argument, since in most cases, axis labels are set automatically.

type

Plot type resp. geom type. May be one of following: "circle" or "tile" circular or tiled geoms, or "bar" for a bar plot. You may use initial letter only for this argument.

geom.size

size resp. width of the geoms (bar width, line thickness or point size, depending on plot type and function). Note that bar and bin widths mostly need smaller values than dot sizes.

geom.colors

user defined color for geoms. See 'Details' in sjp.grpfrq.

wrap.title

numeric, determines how many chars of the plot title are displayed in one line and when a line break is inserted.

wrap.labels

numeric, determines how many chars of the value, variable or axis labels are displayed in one line and when a line break is inserted.

show.values

logical, whether values should be plotted or not.

show.cronb

logical, if TRUE (default), the cronbach's alpha value for each factor scale will be calculated, i.e. all variables with the highest loading for a factor are taken for the reliability test. The result is an alpha value for each factor dimension. Only applies when data is a data frame and no prcomp object.

prnt.plot

logical, if TRUE (default), plots the results as graph. Use FALSE if you don't want to plot any graphs. In either case, the ggplot-object will be returned as value.

Value

(Invisibly) returns a structure with

• the varimax-rotated factor loading matrix (varim)
• the column indices of removed variables (for more details see next list item) (removed.colindex)
• an updated data frame containing all factors that have a clear loading on a specific scale in case data was a data frame (See argument fctr.load.tlrn for more details) (removed.df)
• the factor.index, i.e. the column index of each variable with the highest factor loading for each factor,
• the ggplot-object (plot),
• the data frame that was used for setting up the ggplot-object (df).

See Also

• sjPlot manual: sjp.pca
• sjt.pca

Examples

# randomly create data frame with 7 items, each consisting of 4 categories
likert_4 <- data.frame(
  sample(1:4, 500, replace = TRUE, prob = c(0.2, 0.3, 0.1, 0.4)),
  sample(1:4, 500, replace = TRUE, prob = c(0.5, 0.25, 0.15, 0.1)),
  sample(1:4, 500, replace = TRUE, prob = c(0.4, 0.15, 0.25, 0.2)),
  sample(1:4, 500, replace = TRUE, prob = c(0.25, 0.1, 0.4, 0.25)),
  sample(1:4, 500, replace = TRUE, prob = c(0.1, 0.4, 0.4, 0.1)),
  sample(1:4, 500, replace = TRUE),
  sample(1:4, 500, replace = TRUE, prob = c(0.35, 0.25, 0.15, 0.25))
)

# Create variable labels
colnames(likert_4) <- c("V1", "V2", "V3", "V4", "V5", "V6", "V7")

# plot results from PCA as square-tiled "heatmap"
sjp.pca(likert_4, type = "tile")

# plot results from PCA as bars
sjp.pca(likert_4, type = "bar")

# manually compute PCA
pca <- prcomp(na.omit(likert_4), retx = TRUE, center = TRUE, scale. = TRUE)
# plot results from PCA as circles, including Eigenvalue-diagnostic.
# note that this plot does not compute the Cronbach's Alpha
sjp.pca(pca, plot.eigen = TRUE, type = "circle", geom.size = 10)

# -------------------------------
# Data from the EUROFAMCARE sample dataset
# -------------------------------
library(sjmisc)
data(efc)

# retrieve variable and value labels
varlabs <- get_label(efc)

# recveive first item of COPE-index scale
start <- which(colnames(efc) == "c82cop1")
# recveive last item of COPE-index scale
end <- which(colnames(efc) == "c90cop9")
 
# create data frame with COPE-index scale
mydf <- data.frame(efc[, c(start:end)])
colnames(mydf) <- varlabs[c(start:end)]

sjp.pca(mydf)
sjp.pca(mydf, type = "tile")

# -------------------------------
# auto-detection of labels
# -------------------------------
sjp.pca(efc[, c(start:end)], type = "circle", geom.size = 10)