draw_network example and title family

Author: Rmomal

DESCRIPTION

@@ -22,9 +22,7 @@ Imports:
     tidygraph,
     tidyr,
     vegan,
-    viridisLite,
     ade4,
-    viridis,
     gridExtra
 Suggests: 
     testthat (>= 2.1.0),

NAMESPACE

@@ -77,4 +77,3 @@ importFrom(tidygraph,centrality_betweenness)
 importFrom(tidygraph,centrality_degree)
 importFrom(tidygraph,edge_is_incident)
 importFrom(vegan,spantree)
-importFrom(viridisLite,viridis)

R/F_Graphs.R

@@ -19,25 +19,33 @@
 #' @param node_groups optional vector separating the nodes into groups
 #' @param edge_groups optional matrix specifying groups of edges
 #' @param legend optional boolean for generating a legend when groups are provided
-#' @param pos optional legend position ("bottom", "top", "right" or "left")
+#' @param leg.pos optional legend position ("bottom", "top", "right" or "left")
+#' @param title.family font family of the title
+#' @param title.size font size of the title
 #' @return \itemize{
 #' \item{G} {the network as a ggplot2 object, with highlighted high betweenness nodes}
 #' \item{graph_data}{ data needed for plotting the network}
+#' \item{legend}{ ggplot2 information about the legend}
 #' }
 #' @export
 #' @importFrom tidygraph .N as_tbl_graph activate centrality_betweenness centrality_degree edge_is_incident
 #' @importFrom ggraph ggraph set_graph_style geom_edge_arc scale_edge_alpha_manual scale_edge_width_continuous geom_node_text geom_node_point  scale_edge_colour_manual create_layout
 #' @importFrom tibble tibble rownames_to_column
 #' @importFrom ggplot2 aes theme labs scale_color_manual scale_size_manual ggplot_gtable ggplot_build element_text
 #' @importFrom dplyr mutate filter
-#' @importFrom viridisLite viridis
 #' @importFrom tibble as_tibble
 #' @importFrom gridExtra grid.arrange
-#' @examples adj_matrix= SimCluster(20,2,0.4, 10)
-#' draw_network(adj_matrix,"Cluster graph", layout="stress", shade=TRUE)
+#' @examples
+#' set.seed(1)
+#' adj_matrix= SimCluster(p=20,k=2,dens=0.5, r=10)
+#' groups=sample(3,20, replace=TRUE)
+#' draw_network(adj_matrix,"Cluster graph", layout="stress", shade=TRUE, btw_rank=3)$G
+#' draw_network(adj_matrix,"Cluster with groups of nodes", layout="stress", shade=TRUE, btw_rank=3,
+#' node_groups=groups, legend=TRUE,title.family="mono",title.size=12)$G
 draw_network<-function(adj_matrix,title="", label_size=4, curv=0,width=1, shade=FALSE, remove_isolated=FALSE,btw_rank=2,
                        layout=NULL,stored_layout=NULL,nodes_label=NULL,nodes_size=c(2,5),pal_edges=NULL, pal_nodes=NULL,
-                       node_groups=NULL, edge_groups=NULL,legend=FALSE, pos="bottom"){
+                       node_groups=NULL, edge_groups=NULL,legend=FALSE, leg.pos="bottom",
+                       title.family="sans",title.size=12){
   adj_matrix=as.matrix(adj_matrix)
   p=nrow(adj_matrix) ; binary=FALSE
 
@@ -53,7 +61,11 @@ draw_network<-function(adj_matrix,title="", label_size=4, curv=0,width=1, shade=
   min.width=ifelse(binary,0,0.1)
   #edges colour
 if(is.null(pal_edges)){
-  pal_edges=viridisLite::viridis(5, option = "C")[c(3,2,4,1)][1:length(unique(edge_groups))]
+  ncol_e=length(unique(edge_groups))
+  if(ncol_e<5){
+    pal_edges=c("#31374f","#1F968BFF","#B8DE29FF","de7065ff")[1:ncol_e]
+  }else{stop("Please fill the pal_edges parameter")}
+
 }
   #betweenness computations
 
@@ -90,8 +102,8 @@ if(is.null(pal_edges)){
   }
 
   #draw graph
-  set_graph_style(family="sans")
-  layout = ifelse(is.null(layout), "circle", layout)
+  set_graph_style(title.family)
+  layout = ifelse(is.null(layout), "nicely", layout)
 
   if(is.null(stored_layout)){
     finallayout=create_layout(res,layout=layout)
@@ -120,20 +132,21 @@ if(is.null(pal_edges)){
     geom_node_text(aes(label = label), color = "black", size = label_size) +#,nudge_x = 0.3
     labs(title = title) + theme(plot.title = ggplot2::element_text(
       hjust=0.5,
-      family = "Helvetica",
-      size = 12 ))+
+      family = title.family,
+      size = title.size ),
+      legend.position=leg.pos)+
     scale_edge_width_continuous(range=c(min.width,width))
 leg=NULL
   if(!is.null(node_groups) & legend ){#add legend if groups provided
     tmp=ggplot(data.frame(node_groups=as.factor(node_groups), row1=adj_matrix[1,], row2=adj_matrix[2,]),
                aes(row1, row2, color=node_groups))+
-      geom_point()+scale_color_manual("",values=pal_nodes)+theme(legend.position=pos)
+      geom_point()+scale_color_manual("",values=pal_nodes)+theme(legend.position=leg.pos)
     tmp <- ggplot_gtable(ggplot_build(tmp))
     leg <- tmp$grobs[[which(sapply(tmp$grobs, function(x) x$name) == "guide-box")]]
 
   }
 
-  return(list(G=g,legend=leg,graph_data=res))
+  return(list(G=g,graph_data=res,legend=leg))
 }
 
 utils::globalVariables(c("edges", "weight", "nodes", "btw.weights", "btw",

R/F_inference.R

@@ -141,11 +141,13 @@ Psi_alpha <- function(CorY, n, cond.tol=1e-10){
 #' @importFrom tibble tibble rowid_to_column
 #' @importFrom ggplot2 ggplot geom_point geom_line theme_minimal labs aes
 #' @importFrom stats optim
-#' @examples n=30
+#' @examples
+#' set.seed(1)
+#' n=50
 #' p=10
-#' S=5
 #' Y=data_from_scratch("tree",p=p,n=n)$data
-#' beta = matrix(1/10,10,10)
+#' mean.val=exp((-(p-2)*log(p))/(p-1))
+#' beta = matrix(mean.val, p, p);  diag(beta)=0
 #' psi=Psi_alpha(cor(Y), n)$psi
 #' FitEM = FitBeta(beta.init=beta, psi=psi, maxIter = 6, sum.weights=sum(beta))
 FitBeta <- function(beta.init, psi, maxIter=50, eps = 1e-6, unlinked=NULL,sum.weights){
@@ -394,7 +396,8 @@ StATS<-function(Pmat, nlambda=100, stab.thresh=0.9,plot=FALSE){
 #'}
 #'custom_resample=ResampleEMtree(simu$data,S=S,cores = 1,user_covariance_estimation=estimSigma)
 #'
-#'# We then run the stability selection to find the optimal selection frequencies, for a stability of 85%:
+#'# We then run the stability selection to find the optimal selection frequencies,
+#'# for a stability of 85%:
 #'stab_default=StATS(default_resample$Pmat, nlambda=50, stab.thresh=0.85,plot=TRUE)
 #'stab_custom=StATS(custom_resample$Pmat, nlambda=50, stab.thresh=0.85,plot=TRUE)
 #'

README.Rmd

@@ -17,7 +17,9 @@ knitr::opts_chunk$set(
 [![Codecov test coverage](https://codecov.io/gh/Rmomal/EMtree/branch/master/graph/badge.svg)](https://codecov.io/gh/Rmomal/EMtree?branch=master)
 [![DOI](https://zenodo.org/badge/166967948.svg)](https://zenodo.org/badge/latestdoi/166967948)
 
-> `EMtree` infers direct species association networks, implementing the procedure described in [Momal *et al.*](https://besjournals.onlinelibrary.wiley.com/doi/abs/10.1111/2041-210X.13380). This package uses averages over spanning trees within an Expectation-Maximization algorithm to infer conditional dependence networks, and involves plotting funcitonalities (using `ggraph` and `tydigraph`). By default, it uses the Poisson log-Normal Model ([PLNmodels](https://github.com/jchiquet/PLNmodels>)) to accomodate abundance data. However, EMtree is an inference procedure which only requires an estimate of a Gaussian covariance matrix, and can be used with any model which either use Gaussian latent variables, Gaussian copulas, or Gaussian data transformations.
+> `EMtree` infers direct species association networks, implementing the procedure described in [Momal *et al.*](https://besjournals.onlinelibrary.wiley.com/doi/abs/10.1111/2041-210X.13380). This package uses averages over spanning trees within an Expectation-Maximization algorithm to infer conditional dependence networks, and involves plotting funcitonalities (using `ggraph` and `tydigraph`). 
+
+By default, it uses the Poisson log-Normal Model ([PLNmodels](https://github.com/jchiquet/PLNmodels>)) to accommodate abundance data. However, EMtree is an inference method which only requires an estimate of a Gaussian covariance matrix, and can be used with any model which either use Gaussian latent variables, Gaussian copulas, or Gaussian data transformations.
 
 ## Installation
 
@@ -28,7 +30,7 @@ knitr::opts_chunk$set(
 ```{r,eval=FALSE}
 required_CRAN <- c("Matrix", "purrr","parallel",  "mvtnorm", "vegan","huge",
                    "ggplot2", "magrittr", "dplyr","tidyr", "tibble",
-                   "PLNmodels","ggraph", "tidygraph", "ade4", "viridisLite")
+                   "PLNmodels","ggraph", "tidygraph", "ade4")
 not_installed_CRAN <- setdiff(required_CRAN, rownames(installed.packages()))
 if (length(not_installed_CRAN) > 0) install.packages(not_installed_CRAN)
 ```

README.md

@@ -15,12 +15,14 @@ coverage](https://codecov.io/gh/Rmomal/EMtree/branch/master/graph/badge.svg)](ht
 > This package uses averages over spanning trees within an
 > Expectation-Maximization algorithm to infer conditional dependence
 > networks, and involves plotting funcitonalities (using `ggraph` and
-> `tydigraph`). By default, it uses the Poisson log-Normal Model
-> ([PLNmodels](https://github.com/jchiquet/PLNmodels%3E)) to accomodate
-> abundance data. However, EMtree is an inference procedure which only
-> requires an estimate of a Gaussian covariance matrix, and can be used
-> with any model which either use Gaussian latent variables, Gaussian
-> copulas, or Gaussian data transformations.
+> `tydigraph`).
+
+By default, it uses the Poisson log-Normal Model
+([PLNmodels](https://github.com/jchiquet/PLNmodels%3E)) to accommodate
+abundance data. However, EMtree is an inference method which only
+requires an estimate of a Gaussian covariance matrix, and can be used
+with any model which either use Gaussian latent variables, Gaussian
+copulas, or Gaussian data transformations.
 
 ## Installation
 
@@ -31,7 +33,7 @@ coverage](https://codecov.io/gh/Rmomal/EMtree/branch/master/graph/badge.svg)](ht
 ``` r
 required_CRAN <- c("Matrix", "purrr","parallel",  "mvtnorm", "vegan","huge",
                    "ggplot2", "magrittr", "dplyr","tidyr", "tibble",
-                   "PLNmodels","ggraph", "tidygraph", "ade4", "viridisLite")
+                   "PLNmodels","ggraph", "tidygraph", "ade4")
 not_installed_CRAN <- setdiff(required_CRAN, rownames(installed.packages()))
 if (length(not_installed_CRAN) > 0) install.packages(not_installed_CRAN)
 ```