plot_simtheory: Plot Simulated Data and Target Distribution Data by Name or...

In SimCorrMix: Simulation of Correlated Data with Multiple Variable Types Including Continuous and Count Mixture Distributions

Description

This plots simulated continuous or count (regular or zero-inflated, Poisson or Negative Binomial) data and overlays data (if overlay = TRUE) generated from the target distribution. The target is specified by name (plus up to 4 parameters) or PDF function fx (plus support bounds). Due to the integration involved in finding the CDF from the PDF supplied by fx, only continuous fx may be supplied. Both are plotted as histograms (using geom_histogram). If a continuous target distribution is specified (cont_var = TRUE), the simulated data y is scaled and then transformed (i.e. y = sigma * scale(y) + mu) so that it has the same mean (mu) and variance (sigma^2) as the target distribution. It works for valid or invalid power method PDF's. It returns a ggplot2-package object so the user can save it or modify it as necessary. The graph parameters (i.e. title, sim_color, target_color, legend.position, legend.justification, legend.text.size, title.text.size, axis.text.size, and axis.title.size) are inputs to the ggplot2-package functions so information about valid inputs can be obtained from that package's documentation.

Usage

plot_simtheory(sim_y, title = "Simulated Data Values", ylower = NULL,
  yupper = NULL, sim_color = "dark blue", overlay = TRUE,
  cont_var = TRUE, target_color = "dark green", binwidth = NULL,
  nbins = 100, Dist = c("Benini", "Beta", "Beta-Normal",
  "Birnbaum-Saunders", "Chisq", "Dagum", "Exponential", "Exp-Geometric",
  "Exp-Logarithmic", "Exp-Poisson", "F", "Fisk", "Frechet", "Gamma", "Gaussian",
  "Gompertz", "Gumbel", "Kumaraswamy", "Laplace", "Lindley", "Logistic",
  "Loggamma", "Lognormal", "Lomax", "Makeham", "Maxwell", "Nakagami",
  "Paralogistic", "Pareto", "Perks", "Rayleigh", "Rice", "Singh-Maddala",
  "Skewnormal", "t", "Topp-Leone", "Triangular", "Uniform", "Weibull",
  "Poisson", "Negative_Binomial"), params = NULL, fx = NULL, lower = NULL,
  upper = NULL, seed = 1234, sub = 1000, legend.position = c(0.975,
  0.9), legend.justification = c(1, 1), legend.text.size = 10,
  title.text.size = 15, axis.text.size = 10, axis.title.size = 13)

Arguments

sim_y	a vector of simulated data
title	the title for the graph (default = "Simulated Data Values")
ylower	the lower y value to use in the plot (default = NULL, uses minimum simulated y value) on the y-axis
yupper	the upper y value (default = NULL, uses maximum simulated y value) on the y-axis
sim_color	the histogram fill color for the simulated variable (default = "dark blue")
overlay	if TRUE (default), the target distribution is also plotted given either a distribution name (and parameters) or PDF function fx (with support bounds = lower, upper)
cont_var	TRUE (default) for continuous variables, FALSE for count variables
target_color	the histogram fill color for the target distribution (default = "dark green")
binwidth	the width of bins to use when creating the histograms (default = NULL)
nbins	the number of bins to use when creating the histograms (default = 100); overridden by binwidth
Dist	name of the distribution. The possible values are: "Benini", "Beta", "Beta-Normal", "Birnbaum-Saunders", "Chisq", "Exponential", "Exp-Geometric", "Exp-Logarithmic", "Exp-Poisson", "F", "Fisk", "Frechet", "Gamma", "Gaussian", "Gompertz", "Gumbel", "Kumaraswamy", "Laplace", "Lindley", "Logistic", "Loggamma", "Lognormal", "Lomax", "Makeham", "Maxwell", "Nakagami", "Paralogistic", "Pareto", "Perks", "Rayleigh", "Rice", "Singh-Maddala", "Skewnormal", "t", "Topp-Leone", "Triangular", "Uniform", "Weibull", "Poisson", and "Negative_Binomial". Please refer to the documentation for each package (either stats-package, VGAM-package, or triangle) for information on appropriate parameter inputs.
params	a vector of parameters (up to 4) for the desired distribution (keep NULL if fx supplied instead); for Poisson variables, must be lambda (mean) and the probability of a structural zero (use 0 for regular Poisson variables); for Negative Binomial variables, must be size, mean and the probability of a structural zero (use 0 for regular NB variables)
fx	a PDF input as a function of x only, i.e. fx = function(x) 0.5 * (x - 1)^2; must return a scalar (keep NULL if Dist supplied instead)
lower	the lower support bound for a supplied fx, else keep NULL (note: if an error is thrown from uniroot, try a slightly higher lower bound; i.e., 0.0001 instead of 0)
upper	the upper support bound for a supplied fx, else keep NULL (note: if an error is thrown from uniroot, try a lower upper bound; i.e., 100000 instead of Inf)
seed	the seed value for random number generation (default = 1234)
sub	the number of subdivisions to use in the integration to calculate the CDF from fx; if no result, try increasing sub (requires longer computation time; default = 1000)
legend.position	the position of the legend
legend.justification	the justification of the legend
legend.text.size	the size of the legend labels
title.text.size	the size of the plot title
axis.text.size	the size of the axes text (tick labels)
axis.title.size	the size of the axes titles

Value

A ggplot2-package object.

References

Carnell R (2017). triangle: Provides the Standard Distribution Functions for the Triangle Distribution. R package version 0.11. https://CRAN.R-project.org/package=triangle.

Fialkowski AC (2018). SimMultiCorrData: Simulation of Correlated Data with Multiple Variable Types. R package version 0.2.2. https://CRAN.R-project.org/package=SimMultiCorrData.

Headrick TC, Sheng Y, & Hodis FA (2007). Numerical Computing and Graphics for the Power Method Transformation Using Mathematica. Journal of Statistical Software, 19(3):1-17.
doi: 10.18637/jss.v019.i03.

Wickham H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2009.

Yee TW (2018). VGAM: Vector Generalized Linear and Additive Models. R package version 1.0-5. https://CRAN.R-project.org/package=VGAM.

See Also

calc_theory, ggplot, geom_histogram

Examples

# Using normal mixture variable from contmixvar1 example
Nmix <- contmixvar1(n = 1000, "Polynomial", means = 0, vars = 1,
  mix_pis = c(0.4, 0.6), mix_mus = c(-2, 2), mix_sigmas = c(1, 1),
  mix_skews = c(0, 0), mix_skurts = c(0, 0), mix_fifths = c(0, 0),
  mix_sixths = c(0, 0))
plot_simtheory(Nmix$Y_mix[, 1], title = "Mixture of Normal Distributions",
  fx = function(x) 0.4 * dnorm(x, -2, 1) + 0.6 * dnorm(x, 2, 1),
  lower = -5, upper = 5)
## Not run: 
# Mixture of Beta(6, 3), Beta(4, 1.5), and Beta(10, 20)
Stcum1 <- calc_theory("Beta", c(6, 3))
Stcum2 <- calc_theory("Beta", c(4, 1.5))
Stcum3 <- calc_theory("Beta", c(10, 20))
mix_pis <- c(0.5, 0.2, 0.3)
mix_mus <- c(Stcum1[1], Stcum2[1], Stcum3[1])
mix_sigmas <- c(Stcum1[2], Stcum2[2], Stcum3[2])
mix_skews <- c(Stcum1[3], Stcum2[3], Stcum3[3])
mix_skurts <- c(Stcum1[4], Stcum2[4], Stcum3[4])
mix_fifths <- c(Stcum1[5], Stcum2[5], Stcum3[5])
mix_sixths <- c(Stcum1[6], Stcum2[6], Stcum3[6])
mix_Six <- list(seq(0.01, 10, 0.01), c(0.01, 0.02, 0.03),
  seq(0.01, 10, 0.01))
Bstcum <- calc_mixmoments(mix_pis, mix_mus, mix_sigmas, mix_skews,
  mix_skurts, mix_fifths, mix_sixths)
Bmix <- contmixvar1(n = 10000, "Polynomial", Bstcum[1], Bstcum[2]^2,
  mix_pis, mix_mus, mix_sigmas, mix_skews, mix_skurts, mix_fifths,
  mix_sixths, mix_Six)
plot_simtheory(Bmix$Y_mix[, 1], title = "Mixture of Beta Distributions",
  fx = function(x) mix_pis[1] * dbeta(x, 6, 3) + mix_pis[2] *
    dbeta(x, 4, 1.5) + mix_pis[3] * dbeta(x, 10, 20), lower = 0, upper = 1)

## End(Not run)

SimCorrMix documentation built on May 2, 2019, 1:24 p.m.