Title:	'shiny' Application to Use the Stochastic Block Model
Version:	0.1.5
Description:	A 'shiny' interface for a simpler use of the 'sbm' R package. It also contains useful functions to easily explore the 'sbm' package results. With this package you should be able to use the stochastic block model without any knowledge in R, get automatic reports and nice visuals, as well as learning the basic functions of 'sbm'.
BugReports:	https://github.com/Jo-Theo/shinySbm/issues
License:	MIT + file LICENSE
Depends:	R (≥ 3.50), sbm
Imports:	colourpicker, config (≥ 0.3.1), data.table, dplyr, DT, flextable, fresh, ggplot2, golem (≥ 0.3.5), magrittr, parallel, patchwork, purrr, rmarkdown, shiny (≥ 1.7.2), shinyalert, shinydashboard, stringr, visNetwork
Suggests:	knitr, spelling, testthat (≥ 3.0.0)
Config/testthat/edition:	3
Encoding:	UTF-8
Language:	en-US
LazyData:	true
RoxygenNote:	7.2.3
NeedsCompilation:	no
VignetteBuilder:	knitr
Packaged:	2023-09-07 16:37:53 UTC; theodore
Author:	Theodore Vanrenterghem [cre, aut], Julie Aubert [aut], Saint-Clair Chabert-Liddell [aut], großBM team [ctb], Golem User [cph]
Maintainer:	Theodore Vanrenterghem <shiny.sbm.dev@gmail.com>
Repository:	CRAN
Date/Publication:	2023-09-07 21:50:02 UTC

Pipe operator

Description

See magrittr::%>% for details.

Usage

lhs %>% rhs

Arguments

Value

The result of calling 'rhs(lhs)'.

FungusTreeNetwork

Description

fungus-tree interaction network

This data set provides information about $154$ fungi sampled on $51$ tree species. Composed of nodes and edges lists build based on 'sbm' data package.

Usage

FungusTreeNetwork

Format

A list of the following entries:

networks

• tree_names: (character) tree names

• fungus_names: (character) fungus names

• tree_tree

  1. nodes: data.frame describing nodes of tree_tree network

  2. edges: data.frame describing edges of tree_tree network

  3. type: this network is "unipartite"

• fungus_tree

  1. nodes: data.frame describing nodes of fungus_tree network

  2. edges: data.frame describing edges of fungus_tree network

  3. type: this network is "bipartite"

sbmResults

• tree_treeResults of estimateSimpleSBM for sbm applied on tree_tree data with a Poisson model.

• fungus_treeResults of estimateBipartiteSBM for sbm applied on fungus_tree data with a Bernoulli model.

Source

Vacher, Corinne, Dominique Piou, and Marie-Laure Desprez-Loustau. "Architecture of an antagonistic tree/fungus network: the asymmetric influence of past evolutionary history." PloS one 3.3 (2008): e1740.

get_adjacency

Description

A fct that build an adjacency matrix from a list of edges.

Usage

get_adjacency(edges, type = c("unipartite", "bipartite"), directed = FALSE)

Arguments

Value

an adjacency/incidence matrix (data.frame) representing the network

Examples

# For unipartite network
data_uni <- FungusTreeNetwork$networks$tree_tree

# If the network is symmetric:
my_mat <- get_adjacency(data_uni$edges,
  type = data_uni$type,
  directed = FALSE
)
# If the network is directed:
my_mat <- get_adjacency(data_uni$edges,
  type = data_uni$type,
  directed = TRUE
)

# For bipartite network
data_bi <- FungusTreeNetwork$networks$fungus_tree

my_mat <- get_adjacency(data_bi$edges, type = data_bi$type)

# In any case with a 2 columns data.frames the network is considered binary and each line is a 1.
binary_net <- FungusTreeNetwork$fungus_tree$edges[, -3]

my_mat <- get_adjacency(binary_net, type = data_bi$type)

get_adjacency.default

Description

A fct that build an adjacency matrix from a list of edges

Usage

## Default S3 method:
get_adjacency(edges, type = c("unipartite", "bipartite"), directed = FALSE)

Arguments

Value

an adjacency/incidence matrix (data.frame) representing the network

Examples

# For unipartite network
data_uni <- FungusTreeNetwork$networks$tree_tree

# If the network is symmetric:
my_mat <- get_adjacency(data_uni$edges,
  type = data_uni$type,
  directed = FALSE
)
# If the network is directed:
my_mat <- get_adjacency(data_uni$edges,
  type = data_uni$type,
  directed = TRUE
)

# For bipartite network
data_bi <- FungusTreeNetwork$networks$fungus_tree

my_mat <- get_adjacency(data_bi$edges, type = data_bi$type)

# In any case with a 2 columns data.frames the network is considered binary and each line is a 1.
binary_net <- FungusTreeNetwork$fungus_tree$edges[, -3]

my_mat <- get_adjacency(binary_net, type = data_bi$type)

get_block generic

Description

A fct that return blocks attribution or probabilities for each nodes in a Sbm fit from the sbm package.

Usage

get_block(
  x,
  labels = "default",
  node_names = NULL,
  attribution = TRUE,
  proportion = FALSE
)

Arguments

Value

• "bipartite case": A list containing two data.frames with block attributions and/or proportions one for the row blocks and one for the column blocks

• "unipartite case": A data.frame with block attributions and/or proportions

Examples

# my_sbm_bi <- sbm::estimateBipartiteSBM(sbm::fungusTreeNetwork$fungus_tree,
#                                        model = 'bernoulli')
my_sbm_bi <- FungusTreeNetwork$sbmResults$fungus_tree

node_names_bi <- list(
  row = FungusTreeNetwork$networks$fungus_names,
  col = FungusTreeNetwork$networks$tree_names
)

my_blocks_bi <- get_block(my_sbm_bi,
  labels = c(row = "Fungus", col = "Tree"),
  node_names = node_names_bi
)
my_blocks_bi$row
my_blocks_bi$col

# my_sbm_uni <- sbm::estimateSimpleSBM(sbm::fungusTreeNetwork$tree_tree,
#                                      model = "poisson")
my_sbm_uni <- FungusTreeNetwork$sbmResults$tree_tree

node_names_uni <- list(FungusTreeNetwork$networks$tree_names)

my_blocks_uni <- get_block(my_sbm_uni,
  labels = c("Tree"),
  node_names = node_names_uni
)
my_blocks_uni

get_block.BipartiteSBM_fit method

Description

A fct that return blocks attribution or probabilities for each nodes in a Sbm fit from the sbm package.

Usage

## S3 method for class 'BipartiteSBM_fit'
get_block(
  x,
  labels = "default",
  node_names = NULL,
  attribution = TRUE,
  proportion = FALSE
)

Arguments

Value

A list containing two data.frames with block attributions and/or proportions one for the row blocks and one for the column blocks

Examples

# my_sbm_bi <- sbm::estimateBipartiteSBM(sbm::fungusTreeNetwork$fungus_tree,
#                                        model = 'bernoulli')
my_sbm_bi <- FungusTreeNetwork$sbmResults$fungus_tree

node_names_bi <- list(
  row = FungusTreeNetwork$networks$fungus_names,
  col = FungusTreeNetwork$networks$tree_names
)

my_blocks_bi <- get_block(my_sbm_bi,
  labels = c(row = "Fungus", col = "Tree"),
  node_names = node_names_bi
)
my_blocks_bi$row
my_blocks_bi$col

get_block.SimpleSBM_fit method

Description

A fct that return blocks attribution or probabilities for each nodes in a Sbm fit from the sbm package.

Usage

## S3 method for class 'SimpleSBM_fit'
get_block(
  x,
  labels = "default",
  node_names = NULL,
  attribution = TRUE,
  proportion = FALSE
)

Arguments

Value

A data.frame with block attributions and/or proportions

Examples

# my_sbm_uni <- sbm::estimateSimpleSBM(sbm::fungusTreeNetwork$tree_tree,
#                                      model = "poisson")
my_sbm_uni <- FungusTreeNetwork$sbmResults$tree_tree

node_names_uni <- list(FungusTreeNetwork$networks$tree_names)

my_blocks_uni <- get_block(my_sbm_uni,
  labels = c("Tree"),
  node_names = node_names_uni
)
my_blocks_uni

get_flextable

Description

A fct that build a flextable from an sbm object

Usage

get_flextable(
  sbm,
  labels = "default",
  type = c("blockProp", "connectParam", "storedModels"),
  settings = list()
)

Arguments

Details

Values of type

• 'blockProp': gives the block proportions.

• 'connectParam': gives the block connectivity.

• 'storedModels': gives the stored modems summary.

The list of parameters settings for the flextable

• "caption": Caption is the flextable title (character)

• "digits": nb of digits wanted to be shown in the table

• "selected_col": Color highlighting the selected model

• "best_col": Color of text for the best model

Value

Return the selected flextable

Examples

# my_sbm <- sbm::estimateBipartiteSBM(sbm::fungusTreeNetwork$fungus_tree,
#                                     model = 'bernoulli')
my_sbm <- FungusTreeNetwork$sbmResults$fungus_tree

get_flextable(my_sbm,
  labels = c(row = "Fungus", col = "Trees"),
  type = "blockProp"
)

get_flextable(my_sbm,
  labels = c(row = "Fungus", col = "Trees"),
  type = "connectParam", settings = list(digits = 5)
)

get_flextable(my_sbm,
  labels = "default",
  type = "storedModels", settings = list(caption = "New Title")
)

plotSbm

Description

A fct that plot a beautiful matrix from an sbm object or a network matrix it does have suitable parameters to get the wanted plots. This is the generic function: it does have one method Bipartite and one for Simple Sbm. The 'x' object need to be construct by one of the 'estimate***SBM' function from the 'sbm' package.

Usage

plotSbm(
  x,
  ordered = FALSE,
  transpose = FALSE,
  labels = NULL,
  plotOptions = list()
)

Arguments

Details

The list of parameters plotOptions for the matrix plot is

• "showValues": Boolean. Set TRUE to see the real values. Default value is TRUE

• "showPredictions": Boolean. Set TRUE to see the predicted values. Default value is TRUE

• "title": Title in characters. Will be printed at the bottom of the matrix. Default value is NULL

• "colPred": Color of the predicted values, the small values will be more transparent. Default value is "red"

• "colValue": Color of the real values, the small values will close to white. Default value is "black"

• "showLegend": Should a legend be printed ? TRUE or FALSE, default: FALSE

• "interactionName": Name of connection in legend default: "Connection"

Value

a ggplot object corresponding to the matrix plot inside the app. Groups the network matrix is organized by blocks, the small tiles are for individuals connections. The big tiles between red lines are for block connectivity

Examples

# my_sbm_bi <- sbm::estimateBipartiteSBM(sbm::fungusTreeNetwork$fungus_tree,
#                                        model = 'bernoulli')
my_sbm_bi <- FungusTreeNetwork$sbmResults$fungus_tree

plotSbm(my_sbm_bi,
  ordered = TRUE, transpose = TRUE,
  plotOptions = list(title = "An example Matrix")
)


# my_sbm_uni <- sbm::estimateSimpleSBM(sbm::fungusTreeNetwork$tree_tree,
#                                      model = "poisson")
my_sbm_uni <- FungusTreeNetwork$sbmResults$tree_tree

plotSbm(my_sbm_uni,
  ordered = TRUE,
  plotOptions = list(title = "An example Matrix")
)

n_col <- 100
n_row <- 90
mat <- matrix(sample(0:10, n_col * n_row, replace = TRUE), n_col, n_row)
plotSbm(mat,
  transpose = TRUE,
  labels = list(col = "Columns", row = "Rows"),
  plotOptions = list(colValue = "blue")
)

plotSbm.BipartiteSBM_fit Method

Description

plotSbm method for BipartiteSBM_fit object

Usage

## S3 method for class 'BipartiteSBM_fit'
plotSbm(
  x,
  ordered = FALSE,
  transpose = FALSE,
  labels = NULL,
  plotOptions = list()
)

Arguments

Details

The list of parameters plotOptions for the matrix plot is

• "showValues": Boolean. Set TRUE to see the real values. Default value is TRUE

• "showPredictions": Boolean. Set TRUE to see the predicted values. Default value is TRUE

• "title": Title in characters. Will be printed at the bottom of the matrix. Default value is NULL

• "colPred": Color of the predicted values, the small values will be more transparent. Default value is "red"

• "colValue": Color of the real values, the small values will close to white. Default value is "black"

• "showLegend": Should a legend be printed ? TRUE or FALSE, default: FALSE

• "interactionName": Name of connection in legend default: "Connection"

Value

a ggplot object corresponding to the matrix plot inside the app. Groups the network matrix is organized by blocks, the small tiles are for individuals connections. The big tiles between red lines are for block connectivity

Examples

# my_sbm_bi <- sbm::estimateBipartiteSBM(sbm::fungusTreeNetwork$fungus_tree,
#                                        model = 'bernoulli')
my_sbm_bi <- FungusTreeNetwork$sbmResults$fungus_tree

plotSbm(my_sbm_bi,
  ordered = TRUE, transpose = TRUE,
  plotOptions = list(title = "An example Matrix")
)

plotSbm.SimpleSBM_fit Method

Description

plotSbm method for SimpleSBM_fit object

Usage

## S3 method for class 'SimpleSBM_fit'
plotSbm(
  x,
  ordered = FALSE,
  transpose = FALSE,
  labels = NULL,
  plotOptions = list()
)

Arguments

Details

The list of parameters plotOptions for the matrix plot is

• "showValues": Boolean. Set TRUE to see the real values. Default value is TRUE

• "showPredictions": Boolean. Set TRUE to see the predicted values. Default value is TRUE

• "title": Title in characters. Will be printed at the bottom of the matrix. Default value is NULL

• "colPred": Color of the predicted values, the small values will be more transparent. Default value is "red"

• "colValue": Color of the real values, the small values will close to white. Default value is "black"

• "showLegend": Should a legend be printed ? TRUE or FALSE, default: FALSE

• "interactionName": Name of connection in legend default: "Connection"

Value

a ggplot object corresponding to the matrix plot inside the app. Groups the network matrix is organized by blocks, the small tiles are for individuals connections. The big tiles between red lines are for block connectivity

Examples

# my_sbm_uni <- sbm::estimateSimpleSBM(sbm::fungusTreeNetwork$tree_tree,
#                                      model = "poisson")
my_sbm_uni <- FungusTreeNetwork$sbmResults$tree_tree

plotSbm(my_sbm_uni,
  ordered = TRUE,
  plotOptions = list(title = "An example Matrix")
)

plotSbm.default Method

Description

plotSbm method for unknown object

Usage

## Default S3 method:
plotSbm(
  x,
  ordered = FALSE,
  transpose = FALSE,
  labels = NULL,
  plotOptions = list()
)

Arguments

Value

default plot for x

plotSbm.matrix Method

Description

plotSbm method for matrix object

Usage

## S3 method for class 'matrix'
plotSbm(
  x,
  ordered = FALSE,
  transpose = FALSE,
  labels = NULL,
  plotOptions = list()
)

Arguments

Details

The list of parameters plotOptions for the matrix plot is

• "showValues": Boolean. Set TRUE to see the real values. Default value is TRUE

• "showPredictions": Boolean. Set TRUE to see the predicted values. Default value is TRUE

• "title": Title in characters. Will be printed at the bottom of the matrix. Default value is NULL

• "colPred": Color of the predicted values, the small values will be more transparent. Default value is "red"

• "colValue": Color of the real values, the small values will close to white. Default value is "black"

• "showLegend": Should a legend be printed ? TRUE or FALSE, default: FALSE

• "interactionName": Name of connection in legend default: "Connection"

Value

a ggplot object corresponding to the matrix plot inside the app. Here because there no 'sbm' information and only a matrix describing a network, The matrix isn't organized and the tiles are only showing individuals connections.

Examples

n_col <- 100
n_row <- 90
mat <- matrix(sample(0:10, n_col * n_row, replace = TRUE), n_col, n_row)
plotSbm(mat,
  transpose = TRUE,
  labels = list(col = "Columns", row = "Rows"),
  plotOptions = list(colValue = "blue")
)

Run the Shiny Application

Description

Run the Shiny Application

Usage

shinySbmApp(
  nbCore_control = TRUE,
  console_verbosity = TRUE,
  onStart = NULL,
  options = list(launch.browser = TRUE),
  enableBookmarking = NULL,
  uiPattern = "/",
  ...
)

Arguments

Value

No return value, called to launch the 'shiny' application

visSbm

Description

A fct that plot a visNetwork plot of a adjacency matrix or an Sbm fit from the sbm package.

Usage

visSbm(
  x,
  labels = "default",
  node_names = NULL,
  directed = "default",
  settings = list()
)

Arguments

Details

List of parameters

• "edge_threshold": "default" erases as many small edges as it can without isolating any nodes (no connection). It can also be a numeric value between 0 and 1, relative (between min and max) filter for small edges value

• "edge_color": character: color of edges (default: "lightblue")

• "arrows": boolean: should edges be arrows

• "arrow_thickness": numeric: arrows size

• "arrow_start": character: "row" or "col" or labels value according to row or columns. The arrow will start from selected to the the other value

• "node_color": named character: Bipartite case c(row = "row_color", col = "col_color"). Unipartite case c("node_color")

• "node_shape": named character: Bipartite case c(row = "row_shape", col = "col_shape"). Unipartite case c("node_shape"). Value from visNetwork shape argument of visEdges function ("triangle","dot","square",etc...)

• "digits": integer: number of digits to show when numbers are shown (default: 2)

Value

a visNetwork visual of the x object

Examples

# my_sbm_bi <- sbm::estimateBipartiteSBM(sbm::fungusTreeNetwork$fungus_tree,
#                                        model = 'bernoulli')
my_sbm_bi <- FungusTreeNetwork$sbmResults$fungus_tree

node_names_bi <- list(
  row = FungusTreeNetwork$networks$fungus_names,
  col = FungusTreeNetwork$networks$tree_names
)

visSbm(my_sbm_bi,
  labels = c(row = "Fungus", col = "Tree"),
  node_names = node_names_bi,
  settings = list(
    arrows = TRUE,
    arrow_start = "Fungus",
    node_color = c(row = "pink", col = "green"),
    node_shape = c(row = "dot", col = "square")
  )
)


# my_sbm_uni <- sbm::estimateSimpleSBM(sbm::fungusTreeNetwork$tree_tree,
#                                      model = "poisson")
my_sbm_uni <- FungusTreeNetwork$sbmResults$tree_tree

node_names_uni <- list(FungusTreeNetwork$networks$tree_names)

visSbm(my_sbm_uni,
  labels = c("Tree"),
  node_names = node_names_uni,
  settings = list(
    edge_threshold = 0.01,
    edge_color = "grey",
    node_color = c("violet")
  )
)

visSbm

Description

A fct that plot a visNetwork plot of a adjacency matrix or an Sbm fit from the sbm package.

Usage

## S3 method for class 'BipartiteSBM_fit'
visSbm(
  x,
  labels = "default",
  node_names = NULL,
  directed = "default",
  settings = list()
)

Arguments

Details

List of parameters

• "edge_threshold": "default" erases as many small edges as it can without isolating any nodes (no connection). It can also be a numeric value between 0 and 1, relative (between min and max) filter for small edges value

• "edge_color": character: color of edges (default: "lightblue")

• "arrows": boolean: should edges be arrows

• "arrow_thickness": numeric: arrows size

• "arrow_start": character: "row" or "col" or labels value according to row or columns. The arrow will start from selected to the the other value

• "node_color": named character: Bipartite case c(row = "row_color", col = "col_color"). Unipartite case c("node_color")

• "node_shape": named character: Bipartite case c(row = "row_shape", col = "col_shape"). Unipartite case c("node_shape"). Value from visNetwork shape argument of visEdges function ("triangle","dot","square",etc...)

• "digits": integer: number of digits to show when numbers are shown (default: 2)

Value

a visNetwork visual of the x object

Examples

# my_sbm_bi <- sbm::estimateBipartiteSBM(sbm::fungusTreeNetwork$fungus_tree,
#                                        model = 'bernoulli')
my_sbm_bi <- FungusTreeNetwork$sbmResults$fungus_tree

node_names_bi <- list(
  row = FungusTreeNetwork$networks$fungus_names,
  col = FungusTreeNetwork$networks$tree_names
)

visSbm(my_sbm_bi,
  labels = c(row = "Fungus", col = "Tree"),
  node_names = node_names_bi,
  settings = list(
    arrows = TRUE,
    arrow_start = "Fungus",
    node_color = c(row = "pink", col = "green"),
    node_shape = c(row = "dot", col = "square")
  )
)

visSbm

Description

A fct that plot a visNetwork plot of a adjacency matrix or an Sbm fit from the sbm package.

Usage

## S3 method for class 'SimpleSBM_fit'
visSbm(
  x,
  labels = "default",
  node_names = NULL,
  directed = "default",
  settings = list()
)

Arguments

Details

List of parameters

• "edge_threshold": "default" erases as many small edges as it can without isolating any nodes (no connection). It can also be a numeric value between 0 and 1, relative (between min and max) filter for small edges value

• "edge_color": character: color of edges (default: "lightblue")

• "arrows": boolean: should edges be arrows

• "arrow_thickness": numeric: arrows size

• "arrow_start": character: "row" or "col" or labels value according to row or columns. The arrow will start from selected to the the other value

• "node_color": named character: Bipartite case c(row = "row_color", col = "col_color"). Unipartite case c("node_color")

• "node_shape": named character: Bipartite case c(row = "row_shape", col = "col_shape"). Unipartite case c("node_shape"). Value from visNetwork shape argument of visEdges function ("triangle","dot","square",etc...)

• "digits": integer: number of digits to show when numbers are shown (default: 2)

Value

a visNetwork visual of the x object

Examples

# my_sbm_uni <- sbm::estimateSimpleSBM(sbm::fungusTreeNetwork$tree_tree,
#                                      model = "poisson")
my_sbm_uni <- FungusTreeNetwork$sbmResults$tree_tree

node_names_uni <- list(FungusTreeNetwork$networks$tree_names)

visSbm(my_sbm_uni,
  labels = c("Tree"),
  node_names = node_names_uni,
  settings = list(
    edge_threshold = 0.01,
    edge_color = "grey",
    node_color = c("violet")
  )
)

visSbm

Description

A fct that plot a visNetwork plot of a adjacency matrix or an Sbm fit from the sbm package.

Usage

## Default S3 method:
visSbm(
  x,
  labels = "default",
  node_names = NULL,
  directed = "default",
  settings = list()
)

Arguments

Details

List of parameters

• "edge_threshold": "default" erases as many small edges as it can without isolating any nodes (no connection). It can also be a numeric value between 0 and 1, relative (between min and max) filter for small edges value

• "edge_color": character: color of edges (default: "lightblue")

• "arrows": boolean: should edges be arrows

• "arrow_thickness": numeric: arrows size

• "arrow_start": character: "row" or "col" or labels value according to row or columns. The arrow will start from selected to the the other value

• "node_color": named character: Bipartite case c(row = "row_color", col = "col_color"). Unipartite case c("node_color")

• "node_shape": named character: Bipartite case c(row = "row_shape", col = "col_shape"). Unipartite case c("node_shape"). Value from visNetwork shape argument of visEdges function ("triangle","dot","square",etc...)

• "digits": integer: number of digits to show when numbers are shown (default: 2)

Value

a visNetwork visual of the x object