Package 'sgraph'

Assignment pipe

Description

Pipe an object forward into a function or call expression and update the 'lhs' object with the resulting value. Magrittr imported function, see details and examples in the magrittr package.

Arguments

lhs	An object which serves both as the initial value and as target.
rhs	a function call using the magrittr semantics.

Value

None, used to update the value of lhs.

---

Pipe

Description

Pipe an object forward into a function or call expression. Magrittr imported function, see details and examples in the magrittr package.

Arguments

lhs	A value or the magrittr placeholder.
rhs	A function call using the magrittr semantics.

Value

Result of rhs applied to lhs, see details in magrittr package.

---

Exposition pipe

Description

Expose the names in 'lhs' to the 'rhs' expression. Magrittr imported function, see details and examples in the magrittr package.

Arguments

lhs	A list, environment, or a data.frame.
rhs	An expression where the names in lhs is available.

Value

Result of rhs applied to one or several names of lhs.

---

Modify the edge colors of a sgraph object.

Description

Modify the edge colors of a sgraph object by providing a single color. Also works with a vector of correct size.

Usage

add_edge_color(
  sigma_obj,
  one_color = NULL,
  color_attr = NULL,
  color_palette = "Set2"
)

Arguments

sigma_obj	sgraph object
one_color	A single color to color all of the nodes (hex format)
color_attr	The name of an edge attribute
color_palette	Name of RColorBrewer palette to use

Value

sgraph with modified edge colors

Examples

library(igraph)
library(sgraph)

data(lesMis)

sig <- sigma_from_igraph(igraph = lesMis) %>%
  add_edge_color(one_color = "#ccc")
sig

---

Modify the edge size of a sgraph object.

Description

Modify the edge size of a sgraph by providing a single size

Usage

add_edge_size(sigma_obj, one_size = NULL)

Arguments

sigma_obj	sgraph object
one_size	A single size to use for all edges

Value

sgraph with modified edge sizes

Examples

library(igraph)
library(sgraph)

data(lesMis)

sig <- sigma_from_igraph(igraph = lesMis) %>%
  add_edge_size(one_size = 5)
sig

---

Modify the edge zIndex of a sgraph object.

Description

Modify the edge zIndex

Usage

add_edge_zindex(sigma_obj, zindex)

Arguments

sigma_obj	sgraph object
zindex	Zindex value, larger is drawn above.

Value

sgraph

Examples

library(igraph)
library(sgraph)

data(lesMis)

sig <- sigma_from_igraph(igraph = lesMis) %>%
  add_edge_zindex(zindex = 2)
sig

---

Add nodes information to the igraph object

Description

Modify the node attributes of an existing igraph object by providing a dataframe

Usage

add_igraph_info(igraph, df_nodes, fields = names(df_nodes))

Arguments

igraph	Igraph object to modify
df_nodes	Data frame to add to nodes
fields	Columns of df_nodes to add. First must be the node identifier.

Value

A sgraph object with modified node labels

Examples

library(igraph)
library(sgraph)
data(lesMis)

df_nodes = cbind.data.frame(name = igraph::vertex_attr(lesMis, 'label'),
  log10_degree = degree(lesMis))

igraph = add_igraph_info(lesMis, df_nodes)

sig <- sigma_from_igraph(lesMis) %>%
  add_node_size(size_vector = 'log10_degree')

---

Add a listener

Description

Add a listener to report data from a 'sgraph' object in 'Shiny' back to the R session.

Usage

add_listener(sigma_obj, listener)

Arguments

sigma_obj	Sgraph object, created using the sigma_from_igraph function
listener	Either "clickNode" to listen to node clicks or "hoverNode" to listen to node hover

Value

Sgraph object with listener

---

Modify the node visibility of a sgraph object.

Description

Modify the node hidden attribute of an existing sgraph object. The sgraph R package extends the sigma.js library to enable hidden nodes that will appear upon clicking on their parent group.

Usage

add_node_hidden(sigma_obj, hidden_attr)

Arguments

sigma_obj	sgraph object, returned by sigma_from_igraph function
hidden_attr	Attribute to use to set node hidden value

Value

sgraph object with modified node hidden atribute

---

Modify the node labels of a sgraph object.

Description

Modify the node labels of an existing sgraph object by providing an attribute from the initial igraph to use as the labels.

Usage

add_node_labels(sigma_obj, label_attr = NULL)

Arguments

sigma_obj	sgraph object, returned by sigma_from_igraph function
label_attr	Attribute to use to replace node labels

Value

sgraph object with modified node labels

Examples

library(igraph)
library(sgraph)

data(lesMis)

sig <- sigma_from_igraph(igraph = lesMis) %>%
  add_node_labels(label_attr = 'label')
sig

---

Modify the node size of a sgraph object.

Description

Modify the node size of an existing sgraph object by providing either: (1) A single size to use for all nodes; (2) a vector of node sizes; or (3) a metric to use to scale the nodes.

Usage

add_node_size(
  sigma_obj,
  min_size = 1,
  max_size = 3,
  one_size = NULL,
  size_vector = NULL
)

Arguments

sigma_obj	sgraph object, returned by sigma_from_igraph function
min_size	Minimum node size on the graph (for scaling)
max_size	Maximum node size on the graph (for scaling)
one_size	A single size to use for all nodes
size_vector	An optional vector with the sizes for each node

Value

A sgraph object with modified node sizes

Examples

library(igraph)
library(sgraph)

data(lesMis)

layout <- layout_nicely(lesMis)

# one size for all nodes
sig <- sigma_from_igraph(igraph = lesMis, layout = layout) %>%
  add_node_size(one_size = 3)
sig

# using a vector
custom_size <- log10(degree(lesMis))
sig <- sigma_from_igraph(igraph = lesMis, layout = layout) %>%
 add_node_size(size_vector = custom_size)
sig

---

Convert weights for spring layout

Description

Apply spring weights (revert weights). Can add weak links to selected nodes, but best without.

Usage

convert_to_spring_weights(df_links, selected_nodes = NULL)

Arguments

df_links	Links data frame of a sgraph object
selected_nodes	Nodes identifiers to be used for the selection.

Value

Links data frame of a sgraph object

---

Build a color map

Description

Build a color map

Usage

get_color_map(colors, palette = RColorBrewer::brewer.pal(8, "Dark2"))

Arguments

colors	Groups that will be assigned to colors
palette	Palette to use, typically a RColorBrewer palette

Value

Data frame mapping group names to colors

---

Get the legend for a sgraph network

Description

Get the legend for a sgraph network

Usage

get_legend(colors_map, clusters)

Arguments

colors_map	Color mapping to use, typically built by 'get_color_map'
clusters	Group names

Value

Ggplot object

---

Highlight edges of multiple connected nodes

Description

Using a selection of nodes, highlight edges linking to nodes that are connected to several nodes from the selection. Differentiate multiple connected and fully connected (all selected nodes). Use a maximum number of connected nodes to use lighter colored edges (default 20).

Usage

highlight_multiple_connected(
  df_links,
  selected_nodes,
  n_max = 20,
  dark_cols = c("#ddd", "#444", "#444"),
  light_cols = c("#efefef", "#ddd", "#bbb")
)

Arguments

df_links	Links data frame of a sgraph object
selected_nodes	Nodes identifiers to be used for the selection.
n_max	Maximum number of connected nodes, to use either lighter or darker color sets for edges (default 20).
dark_cols	Three hex values for colors to use with n_max.
light_cols	Three hex values for colors to use with n_max.

Value

Links data frame of a sgraph object

---

Kgraph fit to graph list object

Description

Example function to build a graph list object (list of nodes and links data frames) from a kgraph object (embeddings with cosine similarity cut-off based on random null concept pairs and known related concept pairs)

Usage

kgraph_to_lgraph(l_fit_embeds)

Arguments

l_fit_embeds

kgraph object: embeddings with cosine similarity cut-off based on random null concept pairs and known related concept pairs

Value

graph list object: list of nodes and links data frames

---

Graph list object to igraph object

Description

Build an igraph object from a graph list object (list of nodes and links data frames)

Usage

l_graph_to_igraph(l_graph)

Arguments

l_graph

graph list object: list of nodes and links data frames

Value

igraph object

---

Co-appearances of characters in "Les Miserables"

Description

A graph where the nodes are characters in "Les Miserables" and the edges are times that the characters appeared together in the novel.

Usage

lesMis

Format

An igraph object with 77 nodes and 254 edges

id

numeric id of nodes

label

character label (names) of nodes

value

numeric weight of the edges (number of co-appearances)

Source

D. E. Knuth, The Stanford GraphBase: A Platform for Combinatorial Computing, Addison-Wesley, Reading, MA (1993)

---

Format multiline labels

Description

The sgraph R package extends the sigma.js library to enable multiline labels. The Javascript functions will start new lines on line breaks ('\n') and this function enables to easily format the details of nodes by showing the value of the relationship (edge weights) it has with other nodes. See the Shiny examples for use cases.

Usage

multiline_labels(
  df_nodes,
  display_val_str = "\nP-value: ",
  replace_codes = TRUE,
  label_str = "Label: ",
  group_str = "Group: "
)

Arguments

df_nodes	Nodes data frame of a sgraph object
display_val_str	String that will be prepended to each edge weight
replace_codes	Should the label of the node replace the id ?
label_str	String that will pe prepended to the node label
group_str	String that will pe prepended to the node group

Value

sgraph object with modified node hidden atribute

---

Render a sgraph visualization in Shiny

Description

Render a sgraph visualization in Shiny

Usage

renderSgraph(expr, env = parent.frame(), quoted = FALSE)

Arguments

expr	An expression that creates a sgraph visualization
env	Defaults to parent.frame() (cf. Shiny docs)
quoted	Defaults to FALSE (cf. Shiny docs)

Value

Htmlwidgets render object

---

Scale weights

Description

First try to linearize the weights with the best logarithmic and polynomial, then perform exponential scaling and set upper and lower bounds.

Usage

scale_graph(
  weights,
  exp_scale = exp(1),
  upper_bound_mult = 25,
  lower_bound_const = 5
)

Arguments

weights	Either nodes or links weights vector
exp_scale	Scale for exponential transform
upper_bound_mult	Constant to multiply weights by after scaling. Use to set an upper bound for weights.
lower_bound_const	Constant to set a lower bound for weights. All weights below will be set to lower bound.

Value

Weights vector

---

Build a sgraph object colored by clusters

Description

Wrapper function to build a sigma.js visualization of an igraph object and color it using a 'clusters' attribute. It calls the sigma_from_igraph function, and manages labels, node sizes, color mapping, layouts, and optionally arrows. The clusters attribute can be also be disabled to just use the wrapper to manage the other attributes.

Usage

sgraph_clusters(
  igraph,
  color_map = NULL,
  label = "name",
  clusters = TRUE,
  arrows = FALSE,
  node_size = NULL,
  layout = igraph::layout_with_fr(igraph),
  ...
)

Arguments

igraph	Igraph object
color_map	Use a color mapping to select colors (enables to link the graph with other plots) as returned by the get_color_map function. Leave NULL for automatic colors (default).
label	Name of the igraph attribute to use as labels
clusters	Whether or not to use a column named clusters to color the nodes.
arrows	Whether or not to display arrows on directed edges.
node_size	Passed to add_node_size function, either a numeric or an attribute name. Default is NULL.
layout	Output of an igraph layout (default: layout_with_fr)
...	Passed to sigma_from_igraph

Value

Htmlwidget object

---

Create a UI element for a sgraph visualization in Shiny

Description

Create a UI element for a sgraph visualization in Shiny

Usage

sgraphOutput(outputId, width = "100%", height = "400px")

Arguments

outputId	ID of the UI element
width	Width of the UI element
height	Height of the UI element

Value

Htmlwidgets output object

---

Build a sgraph object from an igraph object

Description

Basic sigma.js visualization of an igraph object, with pipeable syntax.

Usage

sigma_from_igraph(
  igraph,
  layout = NULL,
  label_color = "#fff",
  width = "100%",
  height = "400px",
  elementId = NULL,
  label_grid_cell_size = 200
)

Arguments

igraph	Igraph object
layout	Output of an igraph layout (default: layout_nicely)
label_color	Hex color for labels
width	Width of the output graph (default: fit container)
height	Height of the output graph (default: fit container)
elementId	Do not specify, used by the htmlwidgets package
label_grid_cell_size	Sigma.js corresponding parameter. Roughly goes from 1 to 5000, the smaller the more labels displayed.

Value

Htmlwidget object, meant to be called directly to render a default visualization, or passed to other functions to change attributes (colors, sizes, interactivity, etc.).

Examples

library(sgraph)
data(lesMis)

sig <- sigma_from_igraph(igraph = lesMis)
sig

---