Produces a two-panel figure of [1] the change point distributions as histograms over time and [2] the time series of the fitted topic proportions over time, based on a selected set of change point locations.

pred_gamma_TS_plot produces a time series of the fitted topic proportions over time, based on a selected set of change point locations.

rho_hist: make a plot of the change point distributions as histograms over time.

TS_summary_plot(x, cols = set_TS_summary_plot_cols(), bin_width = 1,
  xname = NULL, border = NA, selection = "median", LDATS = FALSE)

pred_gamma_TS_plot(x, selection = "median", cols = set_gamma_colors(x),
  xname = NULL, together = FALSE, LDATS = FALSE)

rho_hist(x, cols = set_rho_hist_colors(x$rhos), bin_width = 1,
  xname = NULL, border = NA, together = FALSE, LDATS = FALSE)

Arguments

x

Object of class TS_fit produced by TS.

cols

list of elements used to define the colors for the two panels, as generated simply using set_TS_summary_plot_cols. Has two elements rho and gamma, each corresponding to the related panel, and each containing default values for entries named cols, option, and alpha. See set_gamma_colors and set_rho_hist_colors for details on usage.

bin_width

Width of the bins used in the histograms, in units of the x-axis (the time variable used to fit the model).

xname

Label for the x-axis in the summary time series plot. Defaults to NULL, which results in usage of the timename element of the control list (held incontrol$TS_control$timename). To have no label printed, set xname = "".

border

Border for the histogram, default is NA.

selection

Indicator of the change points to use. Currently only defined for "median" and "mode".

LDATS

logical indicating if the plot is part of a larger LDATS plot output.

together

logical indicating if the subplots are part of a larger LDA plot output.

Value

NULL.

Examples

# \donttest{
  data(rodents)
  document_term_table <- rodents$document_term_table
  document_covariate_table <- rodents$document_covariate_table
  LDA_models <- LDA_set(document_term_table, topics = 2)[[1]]
  data <- document_covariate_table
  data$gamma <- LDA_models@gamma
  weights <- document_weights(document_term_table)
  TSmod <- TS(data, gamma ~ 1, nchangepoints = 1, "newmoon", weights)
  TS_summary_plot(TSmod)
  pred_gamma_TS_plot(TSmod)
  rho_hist(TSmod)
# }