plot_alluvial: Alluvial plot

In SubgrPlots: Graphical Displays for Subgroup Analysis in Clinical Trials

Description

this function produces an alluvial diagram. This function is a copy of alluvial::alluvial but a 'rotate' option is added to rotate the labels.

Usage

plot_alluvial(
  ...,
  freq,
  col = "gray",
  border = 0,
  layer,
  hide = FALSE,
  alpha = 0.5,
  gap.width = 0.05,
  xw = 0.1,
  cw = 0.1,
  blocks = TRUE,
  ordering = NULL,
  axis_labels = NULL,
  cex = par("cex"),
  cex.axis = par("cex.axis"),
  rotate = 0,
  las = 1,
  bottom.mar = 2
)

Arguments

...	vectors or data frames, all for the same number of observations
freq	numeric, vector of frequencies of the same length as the number of observations
col	vector of colors of the stripes
border	vector of border colors for the stripes
layer	numeric, order of drawing of the stripes
hide	logical, should particular stripe be plotted
alpha	numeric, vector of transparency of the stripes
gap.width	numeric, relative width of inter-category gaps
xw	numeric, the distance from the set axis to the control points of the xspline
cw	numeric, width of the category axis
blocks	logical, whether to use blocks to tie the flows together at each category, versus contiguous ribbons (also admits character value "bookends")
ordering	list of numeric vectors allowing to reorder the alluvia on each axis separately, see Examples
axis_labels	character, labels of the axes, defaults to variable names in the data
cex	numeric, scaling of fonts of category labels and axis labels respectively. See par.
cex.axis	cex value to be passed to the axis. See help(par)
rotate	angle to rotate the labels. This argument is passed as an srt argument
las	direction of the labels of the covariates. 1 is horizontal, 2 is vertical.
bottom.mar	bottom margin to be passed as a first argument to mar()

Examples

library(dplyr)

# Alluvial plot
# Load the data to be used
data(prca)
dat <- prca
dat$trt = dat$rx
dat %>%
  dplyr::select(trt, bm, hx, pf) %>%
  dplyr::group_by(trt, bm, hx, pf) %>%
  dplyr::summarise(Freq = n()) -> alldat
alldat %>%
  ungroup() %>%
  mutate(trt = ifelse(trt == 0 , "Control", "Treatment"),
         bm = ifelse(bm == 0 , "No", "Yes"),
         hx = ifelse(hx == 0 , "No", "Yes"),
         pf = ifelse(pf == 0 , "No", "Yes"))-> alldat

plot_alluvial(alldat[,c(1,3,2,4)], freq = alldat$Freq,
              xw=0.2,cw = 0.12,cex = 1,
              alpha  = 0.8,
              col=ifelse(alldat$trt == "Treatment","#1f78b4", "#a6cee3"),
              layer = alldat$trt  == 1, rotate = 90)


# Alluvial plot using survival rate at 24 months
data(prca)
dat <- prca
dat %>%
  mutate(survival = factor(ifelse(survtime > 24 , "Yes", "No"), levels = c("No", "Yes")),
         trt = rx) -> dat
dat %>%
  dplyr::select(trt, bm, hx, pf, survival) %>%
  dplyr::group_by(trt, bm, hx, pf, survival) %>%
  dplyr::summarise(Freq = n()) -> alldat
alldat %>%
  ungroup() %>%
  mutate(trt = ifelse(trt == 0 , "Control", "Treatment"),
         bm = ifelse(bm == 0 , "No", "Yes"),
         hx = ifelse(hx == 0 , "No", "Yes")) -> alldat

plot_alluvial(alldat[,c(5,1,3,2,4)], freq = alldat$Freq,
              xw=0.2,cw = 0.12,cex = 1,
              alpha  = 0.8,
              col=ifelse(alldat$survival  == "Yes",
                         ifelse(alldat$trt  == "Treatment","#80b1d3","#d5e2eb"),
                         ifelse(alldat$trt  == "Treatment","#faa8d2","#fbe0ee")),
              layer = alldat$trt  == 1, rotate = 90, las = 2, bottom.mar = 5)

SubgrPlots documentation built on Jan. 29, 2020, 5:07 p.m.