Circular Bar Plot

Description

Function cbarplot can be used to plot 2-dimensional circular bar plots. The circular bar plots can only adopt the height-proportional transformation because of the white space between bars.

Usage

cbarplot(
  x,
  nbins = 36,
  radius = 1/sqrt(base::pi),
  prob = TRUE,
  nlabels = 4,
  col = NULL,
  border = NULL,
  m = NA,
  xlim = NULL,
  ylim = NULL,
  main = NULL
)

Arguments

Value

No return value

Author(s)

Danli Xu <dxu452@aucklanduni.ac.nz>, Yong Wang <yongwang@auckland.ac.nz>

References

Xu, D. and Wang, Y. (2020). Area-proportional Visualization for Circular Data. Journal of Computational and Graphical Statistics, 29, 351-357.

See Also

cdensity, cdotplot, chist

Examples

# 600 observations from two von Mises distributions
library(circular)
x = c(rvonmises(200, circular(pi/4), 5), rvonmises(400, circular(pi), 20))

cbarplot(x)                     
cbarplot(x, prob=FALSE)
cbarplot(x, radius=1, nlabels=0, col="lightblue")
cbarplot(x, radius=1, col="lightblue", border="skyblue4")

Circular Density Curve

Description

Function cdensity can be used to plot 2-dimensional density curves for circular data.

Usage

cdensity(
  f,
  radius = 1/sqrt(base::pi),
  area.prop = TRUE,
  total.area = 1,
  nlabels = 4,
  add = FALSE,
  n = 500,
  col = "red",
  xlim = NULL,
  ylim = NULL,
  main = NULL
)

Arguments

Value

No return value

Author(s)

Danli Xu <dxu452@aucklanduni.ac.nz>, Yong Wang <yongwang@auckland.ac.nz>

References

Xu, D. and Wang, Y. (2020). Area-proportional Visualization for Circular Data. Journal of Computational and Graphical Statistics, 29, 351-357.

See Also

cbarplot, cdotplot, chist

Examples

# 600 observations from two von Mises distributions
library(circular)
x = c(rvonmises(200, circular(pi/4), 5), rvonmises(400, circular(pi), 20))
dvm = function(x, mu=0, kappa=1)   # von Mises density
  exp(kappa * cos(x - mu)) * (2 * pi * besselI(kappa, 0))^(-1)
f = function(x) 1/3 * dvm(x, pi/4, 5) + 2/3 * dvm(x, pi, 20)

cdensity(f)                # plot the density in an area-proportional manner

chist(x)                   # circular histogram
cdensity(f, add=TRUE)      # superimpose the density curve
chist(x, area=FALSE)       # height-proportional circular histogram
cdensity(f, area=FALSE, add=TRUE)   # superimpose the density curve

chist(x, radius=0)                          # rose diagrams
cdensity(f, radius=0, add=TRUE)
chist(x, radius=0, area=FALSE)
cdensity(f, radius=0, area=FALSE, add=TRUE)

Circular Stacked Dot Plot

Description

Function cdotplot can be used to plot 2-dimensional stacked dot plot for circular data.

Usage

cdotplot(
  x,
  nbins = 36,
  radius = 1,
  unit = NA,
  area.prop = TRUE,
  total.area = 1,
  m = NA,
  col = "lightblue",
  border = "skyblue4",
  xlim = NULL,
  ylim = NULL,
  main = NULL,
  x.legend = "bottomright",
  y.legend = NULL
)

Arguments

Details

If the number of observations is relatively small, the usual circular stacked dot plot can be used with unit = 1. If the dataset is large, the dots may become too dense to visualize or count. Setting unit to be any positive integer to allow each dot to represent more than one observation. If the number of observations in one bin is not a multiple of the specified unit, a partial dot can be used to represent the remainder at the top of the bin.

Value

No return value

Author(s)

Danli Xu <dxu452@aucklanduni.ac.nz>, Yong Wang <yongwang@auckland.ac.nz>

References

Xu, D. and Wang, Y. (2020). Area-proportional Visualization for Circular Data. Journal of Computational and Graphical Statistics, 29, 351-357.

See Also

cbarplot, cdensity, chist

Examples

# 30 observations from two von Mises distributions
library(circular)
x = c(rvonmises(10, circular(pi/4), 5), rvonmises(20, circular(pi), 20))
cdotplot(x)                 # area-proportional dot plot
cdotplot(x, area = FALSE)   # height-proportional dot plot

# 900 observations from two von Mises distributions
y = c(rvonmises(300, circular(pi/4), 5), rvonmises(600, circular(pi), 20))
cdotplot(y, nbins=76, unit = 10)      # area-proportional (partial) dot plot 
cdotplot(y, nbins=76, unit = 10, area = FALSE) # height-proportional

Circular Histogram and Rose Diagram

Description

Function chist can be used to plot 2-dimensional histograms and rose diagrams for circular data.

Usage

chist(
  x,
  nbins = 36,
  radius = 1/sqrt(base::pi),
  area.prop = TRUE,
  prob = TRUE,
  total.area = 1,
  nlabels = 4,
  col = "lightblue",
  border = "skyblue4",
  m = NA,
  xlim = NULL,
  ylim = NULL,
  main = NULL
)

Arguments

Value

No return value

Author(s)

Danli Xu <dxu452@aucklanduni.ac.nz>, Yong Wang <yongwang@auckland.ac.nz>

References

Xu, D. and Wang, Y. (2020). Area-proportional Visualization for Circular Data. Journal of Computational and Graphical Statistics, 29, 351-357.

See Also

cbarplot, cdensity, cdotplot

Examples

# 600 observations from two von Mises distributions
library(circular)
x = c(rvonmises(200, circular(pi/4), 5), rvonmises(400, circular(pi), 20))

chist(x)                           # area-proportional circular histgram
chist(x, area = FALSE)             # height-proportional circular histgram
chist(x, radius=0)                 # area-proportional rose diagram
chist(x, radius=0, area=FALSE)     # height-proportional rose diagram

chist(x, prob=FALSE)               # labels for frequency
chist(x, nlabels=0)                # no label
chist(x, xlim=c(-1.7,1))           # use xlim
chist(x, area=FALSE, total=2)      # with scaling
chist(x, area=FALSE, total=NULL)   # without scaling

Circular Transformation Formula

Description

The function performs circular transformation of density or frequency, in an area-proportional or height-proportional manner.

Usage

circtrans(x, radius = 0, area.prop = TRUE, factor = 1)

Arguments

Value

A numerical vector of the transformed values

Author(s)

Danli Xu <dxu452@aucklanduni.ac.nz>, Yong Wang <yongwang@auckland.ac.nz>

References

Xu, D. and Wang, Y. (2020) Area-proportional Visualization for Circular Data. Journal of Computational and Graphical Statistics, 29, 351-357.

See Also

scalefactor

Examples

library(circular)
x = as.vector(rvonmises(20, circular(pi), 10))
circtrans(x)                            # area-proportional transformation
circtrans(x, area.prop = FALSE)         # height-proportional transformation
circtrans(x, factor = 2)                # with a scaling factor

Multi-class Circular Density Curve

Description

Function cmdensity can be used to plot 2-dimensional density curves for circular data with multiple classes. The density curves are stacked to avoid any overlap.

Usage

cmdensity(
  funlist,
  funprop = 1,
  radius = 1/sqrt(base::pi),
  area.prop = TRUE,
  total.area = 1,
  n = 500,
  nlabels = 4,
  cols = NULL,
  borders = NULL,
  xlim = NULL,
  ylim = NULL,
  main = NULL,
  type = c("null", "compass", "clock"),
  add = FALSE,
  x.legend = "bottomright",
  y.legend = NULL,
  fill = TRUE,
  lty = 1,
  lwd = 1
)

Arguments

Value

No return value

Author(s)

Danli Xu <dxu452@aucklanduni.ac.nz>, Yong Wang <yongwang@auckland.ac.nz>

References

Xu, D. and Wang, Y. (2020). Area-proportional Visualization for Circular Data. Journal of Computational and Graphical Statistics, 29, 351-357.

See Also

cdensity, cmhist

Examples

# Load and pre-process the dataset
library(circular)
data("pigeons", package = "circular")
x = pigeons[,2] / 180 * pi                  # bearing
y = pigeons[,1]                             # treatment
vs = split(x, factor(y, unique(y)))    # list of classified value
prop = sapply(vs, length) / length(x)  # proportion of each class

# Define the kde function for each class using von Mises kernels
dvm = function(x, mu=0, kappa=1)  # von Mises density
  exp(kappa * cos(x - mu)) * (2 * pi * besselI(kappa, 0))^(-1)
kdevm = function(x, x0, bw=0.3) 
  rowMeans(outer(x, x0, dvm, 0.5 / (1 - exp(-bw^2 / 2))))
fs = list(function(x) kdevm(x, x0=vs[[1]]),
          function(x) kdevm(x, x0=vs[[2]]),
          function(x) kdevm(x, x0=vs[[3]]))

# stacked density curves for 3 classes
cmdensity(fs)                         # 1:1:1
cmdensity(fs, prop)                   # using proportions for functions

Multi-class Stacked Circular Histogram and Rose Diagram

Description

Function cmhist can be used to plot 2-dimensional histograms and rose diagrams for circular data with multiple classes. The histograms are stacked to avoid any overlap.

Usage

cmhist(
  value,
  class,
  nbins = 36,
  radius = 1/sqrt(base::pi),
  area.prop = TRUE,
  prob = TRUE,
  proportion = FALSE,
  total.area = 1,
  nlabels = 4,
  cols = NULL,
  borders = NULL,
  m = NA,
  xlim = NULL,
  ylim = NULL,
  main = NULL,
  type = c("null", "compass", "clock"),
  x.legend = "bottomright",
  y.legend = NULL
)

Arguments

Value

No return value

Author(s)

Danli Xu <dxu452@aucklanduni.ac.nz>, Yong Wang <yongwang@auckland.ac.nz>

References

Xu, D. and Wang, Y. (2020). Area-proportional Visualization for Circular Data. Journal of Computational and Graphical Statistics, 29, 351-357.

See Also

chist, cmdensity

Examples

# Load the dataset
library(circular)
data("pigeons", package = "circular")
x = pigeons[,2] / 180 * pi
y = pigeons[,1]

# stacked circular histograms
cmhist(x, y)             # area-proportional 
cmhist(x, y, area=FALSE) # height-proportional

Scaling Factor

Description

The function calculates the scaling factor so that after scaling the original density curve (before transformation), the total area after transformation (excluding the reference circle) has the specified value.

Usage

scalefactor(x, radius = 0, total.area = 1, area.prop = TRUE)

Arguments

Details

Each value in x is a density value before transformation, for points equally-spaced on [0,2\pi). For a smooth density curve, use a reasonably large number of points, equally-spaced on [0,2\pi). The area under the density curve after transformation is then approximated by that of the corresponding sectors. Note if area.prop = TRUE, the scale factor is simply the value of total.area.

Value

A numerical value for the scaling factor

Author(s)

Danli Xu <dxu452@aucklanduni.ac.nz>, Yong Wang <yongwang@auckland.ac.nz>

References

Xu, D. and Wang, Y. (2020). Area-proportional Visualization for Circular Data. Journal of Computational and Graphical Statistics, 29, 351-357.

See Also

circtrans

Examples

dvm = function(x, mu=0, kappa=1)   # von Mises density
exp(kappa * cos(x - mu)) * (2 * pi * besselI(kappa, 0))^(-1)
x = dvm(seq(0, 2 * pi, len = 100), pi, 10)

scalefactor(x)                            # area-proportional transformation
scalefactor(x, area.prop = FALSE)         # height-proportional transformation
scalefactor(x, total.area = 2)            # total area of 2
scalefactor(x, area.prop = FALSE, total.area = 2)