gg_plotdist: ggplot empirical and theoretical distributions for...

In Issoufou-Liman/decisionSupportExtra: Miscellaneous Functions for Quantitative Support of Decision Making under Uncertainty

Description

Generate compact plot of Bayesian network node states following the descdist function as customizable ggplot.

Usage

gg_plotdist(
  fitted,
  title = NULL,
  hist_bar_size = 0.3,
  dist_geom_pts_size = 0.03,
  dist_geom_smooth_size = 0.3,
  dist_geom_abline_size = 0.3,
  axis_text_size = 9,
  axis_title_size = 9,
  plot_title_size = 10,
  plot_subtitle_size = 8,
  strip_text_size = 9
)

Arguments

fitted	A list of fitdist object
title	character string to be used as plot title
hist_bar_size	numeric size of histogram bars
dist_geom_pts_size, dist_geom_smooth_size, dist_geom_abline_size	numeric. The size of the geom_point, smoothing line and abline, respectively.
axis_text_size	= 12, Text size respectively corresponding to axis text (default to 12), axis title (default to 12), plot title (default to 20), subtitle (default to 17), strip text (default to 18), and legend (default to 12).
axis_title_size	= 12, Text size respectively corresponding to axis text (default to 12), axis title (default to 12), plot title (default to 20), subtitle (default to 17), strip text (default to 18), and legend (default to 12).
plot_title_size	= 12, Text size respectively corresponding to axis text (default to 12), axis title (default to 12), plot title (default to 20), subtitle (default to 17), strip text (default to 18), and legend (default to 12).
plot_subtitle_size	= 12, Text size respectively corresponding to axis text (default to 12), axis title (default to 12), plot title (default to 20), subtitle (default to 17), strip text (default to 18), and legend (default to 12).
strip_text_size	= 12, Text size respectively corresponding to axis text (default to 12), axis title (default to 12), plot title (default to 20), subtitle (default to 17), strip text (default to 18), and legend (default to 12).

Details

see plotdist.

Author(s)

Issoufou Liman

References

Marie Laure Delignette-Muller, Christophe Dutang (2015). fitdistrplus: An R Package for Fitting Distributions. Journal of Statistical Software, 64(4), 1-34. http://www.jstatsoft.org/v64/i04/.

See Also

plotdist.

Examples

library (gRain)
library(bnlearn)
## setting a bayesian network with gRain
Soil_type <- cptable (~Soil_type, values = c(0.05, 0.55, 0.4),
levels = c('Sandy', 'Loamy', 'Clayey'))
Manure_application <- cptable(~Manure_application, values = c(0.3, 0.7),
levels = c('FALSE', 'TRUE'))
Soil_water_holding_capacity_tmp <- make_gRain_CPT(
 parent_effects = list(c(0, 2.5, 3), c(0, 2)),
 parent_weights = c(2,1),
 b = 3,
 child_prior = c(0.2,0.5,0.3),
 child_states = c('Low', 'Medium', 'High'),
 parent_states = list(c('Sandy', 'Loamy', 'Clayey'), c('FALSE', 'TRUE'))
)
Soil_water_holding_capacity_values <- Soil_water_holding_capacity_tmp$values
Soil_water_holding_capacity_levels <- Soil_water_holding_capacity_tmp$levels
Soil_water_holding_capacity <- cptable (
~Soil_water_holding_capacity|Soil_type:Manure_application,
values = Soil_water_holding_capacity_values,
levels = Soil_water_holding_capacity_levels)
## Compile conditional probability tables
network <- compileCPT(list(Soil_type, Manure_application, Soil_water_holding_capacity))
## Graphical Independence Network ####
network <- grain(network)
## Use grain object (gRain package)
fitted <- fit_node_states_distr (bn = network,
node = "Soil_water_holding_capacity", gof="KS")
gg_plotdist(fitted, title = "Soil_water_holding_capacity")
## converting the grain bayesian network to bn.fit
network_bn_fit <- as.bn.fit(network)
## Use bn.fit object (bnlearn package)
fitted <- fit_node_states_distr (bn = network_bn_fit,
node = "Soil_water_holding_capacity", distr = c("beta", "beta", "beta"))
gg_plotdist(fitted, title = "Soil_water_holding_capacity")

Issoufou-Liman/decisionSupportExtra documentation built on Dec. 21, 2020, 6:28 p.m.