rpv_plot {eiExpand}

Racially Polarized Voting Analysis (RPV) Plot

Description

Creates a custom visualization of RPV results

Usage

rpv_plot(
  rpvDF = NULL,
  title = "Racially Polarized Voting Analysis Results",
  subtitle = "Estimated Vote for Candidates by Race",
  legend_name = "Voters' Race:",
  voter_races = NULL,
  colors = NULL,
  position_dodge_width = 0.8,
  bar_size = NULL,
  label_size = 4,
  contest_name_size = 20,
  cand_name_size = 6,
  contest_name_pad = NULL,
  cand_name_pad = -1.5,
  contest_sep = NULL,
  shade_col = "grey75",
  shade_alpha = 0.1,
  panel_spacing = NULL,
  breaks = seq(0, 100, 20),
  lims = c(0, 110),
  includeErrorBand = FALSE,
  includeCandName = TRUE,
  panelBy = NULL,
  txtInBar = NULL
)

Arguments

rpvDF

A data.frame containing RPV results

title

The plot title

subtitle

The plot subtitle

legend_name

The legend title

voter_races

A vector of the unique voter races contained in the Voter_Race column of rpvDF. This argument will set the order in which voter races are displayed in the plot and legend. Can be used with colors, to indicate the which color of the plot to associate with each voter race.

colors

Defines the plot colors for the voter race groups. Colors must be listed in the desired order with respect voter_races if arguments are used together.

position_dodge_width

The width value indicating spacing between the plot bars. Passed to position_dodge().

bar_size

The size of plot bars. Passed to geom_linerange().

label_size

The size of RPV estimate label

contest_name_size

Text size of contest name

cand_name_size

Text size of candidate names if includeCandName = TRUE

contest_name_pad

Padding between contest name and y axis

cand_name_pad

Padding between candidate name and y axis if includeCandName = TRUE.

contest_sep

String indicating how to separate contest. Options "s", "shade", or "shading" shade the background of every other contest. Options "l", "line", "lines" create light grey double lines between contests.

shade_col

color to shade contest separation bands when contest_sep = "s". Defaults to light grey.

shade_alpha

alpha parameter passed to geom_tile() to indicate transparency of contest separation bands when contest_sep = "s"

panel_spacing

Space between facet grid panels

breaks

Numeric vector containing x axis breaks

lims

Numeric vector containing x axis limits

includeErrorBand

Logical indicating if the confidence interval band should appear on the plot. If TRUE, the RPV estimate labels will appear in the middle of each bar instead of at the ends so they don't cover the error bands.

includeCandName

Logical indicating if candidate names should appear on the left side of the plot.

panelBy

Column name from rpvDF passed to facet_grid() to create panels.

txtInBar

Logical indicating location of the RPV estimate labels. If, TRUE, estimates will be in the middle of the plot bars. If FALSE, they will be at the end of the bars.

Value

Bar plot visualization of RPV analysis as a ggplot2 object

Author(s)

Rachel Carroll <[e-mail]>

Loren Collingwood <[e-mail]>

Kassra Oskooii <[e-mail]>

Examples

library(eiExpand)
data(example_rpvDF)

# Note that these plots are designed to be 
# saved as a png using ggplot2::ggsave(). See first example for recommending 
# sizing, noting that height and weight arguments may need adjusting 
# depending on plot attributes such as number of contests and paneling 

# plot county-level results with all defaults
rpvDF_county <- example_rpvDF %>% dplyr::filter(Jurisdiction == "County")
rpv_plot(rpvDF_county)

# save to png with recommended sizing
# ggplot2::ggsave("rpv_plot_default.png", height = 10, width = 15)

# include CI bands
rpv_plot(rpvDF_county, includeErrorBand = TRUE)

# include CI bands with estimate labels outside bar
rpv_plot(
  rpvDF_county, 
  includeErrorBand = TRUE,
  txtInBar = FALSE
)

# panel by preferred candidate
rpvDF_county$Year <- paste(rpvDF_county$Year,
                           "\n") # so contest and year are on different lines
rpvDF_county$Preferred_Candidate <- paste(rpvDF_county$Preferred_Candidate, 
                                          "\nPreferred Candidate")
rpv_plot(
  rpvDF_county,
  panel_spacing = 6,
  panelBy = "Preferred_Candidate"
)

# plot all jurisdictions with panels
rpv_plot(example_rpvDF, panelBy = "Jurisdiction")
# add contest separation shading
rpv_plot(
  example_rpvDF, 
  panelBy = "Jurisdiction",
  contest_sep = "s"
)

# plot panels by voter_race and remove legend
rpv_plot(rpvDF_county,
panel_spacing = 6, 
panelBy = "Voter_Race") + 
  ggplot2::theme(legend.position="none")