plot.sparseDFM: sparseDFM Plot Outputs

Description

Make plots for the output of sparseDFM(). Options include:

factor - plot factor estimate series on top of the original standardized stationary data
loading.heatmap - make a heatmap of the loadings matrix
loading.lineplot - make a lineplot of variable loadings for a given factor
loading.grouplineplot - separate variable groups into colours for better visualisation
residual - boxplot or scatterplot of residuals
lasso.bic - BIC values for the LASSO tuning parameter
em.convergence - log-likelihood convergence of EM iterations

Usage

# S3 method for sparseDFM
plot(
  x,
  type = "factor",
  which.factors = 1:(dim(x$state$factors)[2]),
  scale.factors = TRUE,
  which.series = 1:(dim(x$params$Lambda)[1]),
  loading.factor = 1,
  series.col = "grey",
  factor.col = "black",
  factor.lwd = 2,
  factor.lab = NULL,
  use.series.names = FALSE,
  series.lab = NULL,
  series.labpos = NULL,
  colorkey = TRUE,
  col.regions = NULL,
  group.names = NULL,
  group.cols = NULL,
  group.legend = TRUE,
  residual.type = "boxplot",
  scatter.series = 1,
  min.bic.col = "red",
  alpha_index = "best",
  ...
)

Value

Plots for the output of sparseDFM().

Arguments

x: an object of class 'sparseDFM'.
type: character. The type of plot: "factor", "loading.heatmap", "loading.lineplot", "loading.grouplineplot" or "residual". Default is "factor".
which.factors: numeric vector of integers representing which factors should be plotted in "factor" and "loading.heatmap". Default is which.factors=1:(dim(x$state$factors)[2]), plotting them all. Accepts a single integer if just one factor required.
scale.factors: logical. Standardize the factor estimates when plotting in "factor". Default is TRUE.
which.series: numeric vector of integers representing which series should be plotted in "loading.heatmap", "loading.lineplot", "loading.grouplineplot" and "residual". Default is which.series = 1:(dim(x$params$Lambda)[1]), plotting them all.
loading.factor: integer. The factor to use in "loading.lineplot" and "loading.grouplineplot". Default is 1.
series.col: character. The colour of the background series plotted in "factor". Default is series.col = "grey".
factor.col: character. The colour of the factor estimate line in "factor". Default is factor.col = "black".
factor.lwd: integer. The line width of the factor estimate line in "factor". Default is factor.lwd = 2.
factor.lab: vector of characters to label each factor in "loading.heatmap". Default is NULL for standard labeling.
use.series.names: logical. Set to TRUE if plot should display series names in the data matrix X. Default is FALSE for numbered series.
series.lab: vector of characters to label each data series in "loading.heatmap". Default is NULL for standard labeling.
series.labpos: numeric vector of integers representing which series are labeled by series.lab. Default is NULL for standard labeling.
colorkey: logical. Display the colour key of the heatmap in "loading.heatmap". Default is TRUE.
col.regions: vector of gradually varying colors for "loading.heatmap", see levelplot package. Default is NULL for standard colours.
group.names: vector of characters of the same dimension as which.series to represent the name of the group for each series in "loading.grouplineplot".
group.cols: vector of characters of the same dimension as the number of different groups in "loading.grouplineplot" to represent the colours of the groups.
group.legend: logical. Display the legend. Default is TRUE.
residual.type: character. The type of residual plot: "boxplot" or "scatterplot". Default is "boxplot".
scatter.series: integer. The series to plot when residual.type = "scatterplot". Default is series 1.
min.bic.col: character. Colour for the best $\alpha$ point. Default is 'red'.
alpha_index: Choose which L1 penalty parameter to display the results for. Default is 'best'. Otherwise, input a number between 1:length(alpha_grid) that indicates the required alpha parameter.
...: for plot.sparseDFM. Further plot arguments.