Skip to contents

Scatterplots, sunflower plots, and lowess smoothing

Source: R/plotcmds.R
tda_pl_scatter.Rd

scplot: three related ways of showing an x/y relationship, all in one command via type – plain symbols (the default), a sunflower plot (symbol density shown as petal count, for data too crowded for plain symbols to read), or a lowess-smoothed curve.

Usage

tda_pl_scatter(
  p,
  x,
  y,
  type = c("points", "sunflower", "lowess"),
  symbol = NULL,
  size = NULL,
  grid = NULL,
  bandwidth = NULL,
  select = NULL,
  clip = TRUE,
  lty = NULL,
  lw = NULL,
  ...
)

Arguments

p

a tda_ps session.

x, y

variables in the session's data, or coordinates.

type

"points" (default), "sunflower", or "lowess".

symbol

marker symbol, for type = "points".

size

marker (or, for type = "sunflower", petal) size in mm.

grid

for type = "sunflower", the grid the plot area is divided into to group nearby points, as c(nx, ny) or one number for both; TDA's default is c(1, 1).

bandwidth

for type = "lowess", the smoothing factor; TDA's default is 0.5.

select

a case-selection expression, TDA's sel=.

clip

clip points to the plot's declared range (the default); FALSE turns this off (scplot's nc=1), letting a point outside the range draw past the frame.

lty, lw

line type and width for type = "lowess"'s fitted curve.

...

further options for the command, e.g. d=/ns= for type = "lowess" – distinct from bandwidth's sig=, and not promoted to their named parameters since TDA's help text for them ("optional parameter for opt=3") does not say what they control.

Value

The session, with the command added.

See also

Other plotting: plot.tda_ple(), tda_check_ps(), tda_pl(), tda_pl_arc(), tda_pl_axis(), tda_pl_graph(), tda_pl_hist, tda_pl_panel(), tda_pl_regression(), tda_plot_ps(), tda_ps(), tda_ps3(), tda_read_ps()

Examples

set.seed(1)
d <- data.frame(x = round(runif(30, 0, 10), 1),
                y = round(runif(30, 0, 10), 1))
p <- tda_ps(d, xlim = c(0, 10), ylim = c(0, 10))
p <- tda_pl_frame(p)
p <- tda_pl_scatter(p, "x", "y", symbol = 1)
plot(p)


# a lowess-smoothed curve instead, on data with a real nonlinear trend
d2 <- data.frame(x = sort(round(runif(30, 0, 10), 1)))
d2$y <- round(0.5 * d2$x^1.5 + rnorm(30, sd = 1), 2)
p2 <- tda_ps(d2, xlim = c(0, 10), ylim = c(0, 20))
p2 <- tda_pl_frame(p2)
p2 <- tda_pl_scatter(p2, "x", "y", type = "lowess", bandwidth = 0.6,
                     lty = "dashed")
plot(p2)