L1centrality 0.4.0

Major updates

summary() methods now return an object of class table.
A new argument weight_trasform is added to L1cent(), L1centGROUP(), L1centNB(), L1centLOC(), L1centMDS(), and group_reduce() generic functions. This is an optional function to transform the edge weights.
plot() methods added and enhanced:
- plot.L1centMDS(): Vertex color and vertex label attributes in an igraph object are automatically passed to this plot method.
- plot.L1cent(): When there is no second argument (y), it draws a Lorenz curve. Otherwise, it draws a scatter plot. Axes labels are automatically added.
- plot.L1centLOC(): When the graph is evaluated at a single locality level (alpha), it draws a Lorenz curve or a scatter plot. Otherwise, it draws a line plot of local L1 centrality change of each vertex at a uniform margin.
- plot.L1centEDGE(): Draws a directed graph where each vertex gives an edge to its local median. Colors and labels of the vertices are automatically set from the given igraph object.

Added citations of three relevant papers to the documentations, DESCRIPTION, and README files.
Generated a CITATION file for citing this package.
Removed ‘Summary’ documentation and incorporated the documentation for each summary() method into the corresponding function help pages.
Minor fix in the print() and summary() methods.
Examples modified.

print.L1centNB() can now handle a graph with a missing vertex name (i.e., V(g)$name == NULL). If vertex names are missing, the function automatically sets them to V1, V2, …, and prints them.
The type argument is set to 1 in the quantile() function inside the L1centLOC() and L1centEDGE() functions.

A new function Gini() computes the Gini coefficient. (See the document named Heterogeneity.)
L1cent() now returns an object of class L1cent. summary() and print() methods are implemented for this new class.
L1centGROUP() now returns an object of class L1centGROUP. print() method is implemented for this new class.
L1centLOC() now returns an object of class L1centLOC. summary() and print() methods are implemented for this new class.
L1centNB() now returns an object of class L1centNB. summary() and print() methods are implemented for this new class.
L1centEDGE() now returns an object of class L1centEDGE. summary() and print() methods are implemented for this new class.
Implementation of L1centEDGE() corrected.
print() methods is implemented for the L1centMDS class. The L1centMDS() function formerly returned a length four list with label of the vertices as one component. This component is now an attribute of the returned list, i.e., the L1centMDS() now returns a length three list. (See the document for L1centMDS().) As a result, the plot() method for the L1centMDS is modified as well.

Minor typo fix in the document.
New document ‘Heterogeneity’ added. It includes the Lorenz_plot() function and the Gini() function.
New document ‘Summary’ added. It includes all summary() method in this package.

A new function group_reduce() constructs the group reduced graph, which is used in computing the group L1 centrality and group L1 prestige.
A new function L1centGROUP() computes the group L1 centrality and group L1 prestige. It supports undirected and directed graphs, as well as vertex and edge weighted graphs.

Error message of validate_matrix() (an internal function for validating the given graph) corrected.
DESCRIPTION and the documentation L1centrality-package updated to include information on the group L1 centrality/prestige.
Minor typo fix in the document.

L1cent(), L1centNB(), L1centLOC() now support (strongly connected) directed graphs. Use the new parameter mode to measures the prominence of each vertex in terms of making a choice (mode = "centrality") or receiving a choice (mode = "prestige"). If the graph is undirected, mode does not affect the outcome.

Minor typo fix in the document.
L1centLOC() and L1centEDGE() made more efficient.
\() changed to function() for compatibility (L1centMDS.matrix()).
The vertex multiplicity parameter eta generalized to nonnegative values (L1cent(), L1centEDGE(), L1centLOC(), L1centNB()). It previously accepted only positive values.

In L1centMDS(), setting verbose=FALSE suppresses all the console messages. In the previous version, the final message was printed even when verbose=FALSE.