bar_impvar {genpathmox}

Bar Plot of a ranking of categorical variables by importance

Description

"bar_impvar" returns a bar plot to visualize the ranking of variables by importance in obtaining the terminal nodes of Pathmox.

Usage

bar_impvar(x, .cex.names = 1, .cex.axis = 1.2, .cex.main = 1, ...)

Arguments

x

An object of the class "plstree"

.cex.names

Expansion factor for axis names (bar labels)

.cex.axis

Expansion factor for numeric axis labels

.cex.main

Allows fixing the size of the main. Equal to 1 to default

...

Further arguments are ignored

Details

The importance of each variable is determined by adding the F-statistic calculated for the variable in each split node of Pathmox.

Author(s)

Giuseppe Lamberti

References

Lamberti, G., Aluja, T. B., and Sanchez, G. (2016). The Pathmox approach for PLS path modeling segmentation. Applied Stochastic Models in Business and Industry, 32(4), 453-468. doi: 10.1002/asmb.2168

Lamberti, G. (2015). Modeling with Heterogeneity, PhD Dissertation.

Sanchez, G. (2009). PATHMOX Approach: Segmentation Trees in Partial Least Squares Path Modeling, PhD Dissertation.

See Also

summary.plstree, print.plstree, pls.pathmox, bar_terminal, and plot.plstree

Examples

 ## Not run: 
# Example of PATHMOX approach in customer satisfaction analysis 
# (Spanish financial company).
# Model with 5 LVs (4 common factor: Image (IMAG), Value (VAL), 
# Satisfaction (SAT), and Loyalty (LOY); and 1 composite construct: 
# Quality (QUAL)

# load library and dataset csibank
library(genpathmx)
data("csibank")

# Define the model using the lavaan syntax. Use a set of regression formulas to define
# first the structural model and then the measurement model

CSImodel <- "
# Structural model
VAL  ~ QUAL
SAT  ~ IMAG  + QUAL + VAL
LOY  ~ IMAG + SAT

# Measurement model
# Composite
QUAL <~ qual1 + qual2 + qual3 + qual4 + qual5 + qual6 + qual7 
     
# Common factor
IMAG =~ imag1 + imag2 + imag3 + imag4 + imag5 + imag6 
VAL  =~ val1  + val2  + val3  + val4
SAT  =~ sat1  + sat2  + sat3           
LOY  =~ loy1  + loy2  + loy3           

"

# Identify the categorical variable to be used as input variables 
in the split process
CSIcatvar = csibank[,1:5]

# Check if variables are well specified (they have to be factors 
# and/or ordered factors)
str(CSIcatvar)

# Transform age and education into ordered factors
CSIcatvar$Age = factor(CSIcatvar$Age, levels = c("<=25", 
                                     "26-35", "36-45", "46-55", 
                                     "56-65", ">=66"),ordered = T)

CSIcatvar$Education = factor(CSIcatvar$Education, 
                            levels = c("Unfinished","Elementary", "Highschool",
                            "Undergrad", "Graduated"),ordered = T)
       
# Run Pathmox analysis (Lamberti et al., 2016; 2017)
csi.pathmox = pls.pathmox(
 .model = CSImodel ,
 .data  = csibank,
 .catvar= CSIcatvar,
 .alpha = 0.05,
 .deep = 2
)                     
 
bar_impvar(csi.pathmox)


## End(Not run)