Merge pull request #86 from snlab-ch/develop

Added as_network and preparation for CRAN submission

Author: jhollway

.Rbuildignore

@@ -7,3 +7,4 @@
 ^data-raw$
 ^docs$
 ^pkgdown$
+^cran-comments\.md$

DESCRIPTION

@@ -1,13 +1,22 @@
 Package: migraph
-Title: Multimodal and multilevel network analysis
-Version: 0.6.0
-Date: 2021-03-03
-Description: This package assembles a number of utility functions for 
-   loading, visualising, and analysing multimode, multiplex, and multilevel networks.
-URL: https://github.com/jhollway/migraph
-BugReports: https://github.com/jhollway/migraph/issues
+Title: Multimodal and Multilevel Network Analysis
+Version: 0.6.1
+Date: 2021-04-11
+Description: A set of tools that extend common social network analysis packages
+   for analysing multimodal and multilevel networks.
+   It includes functions for one- and two-mode (and sometimes three-mode) 
+   centrality, centralization, clustering, and constraint,
+   as well as for one- and two-mode network regression and block-modelling.
+   All functions operate with matrices, edge lists, 
+   and 'igraph', 'network'/'sna', and 'tidygraph' objects.
+   The package is released as a complement to 
+   'Multimodal Political Networks' (2021, ISBN:9781108985000),
+   and includes various datasets used in the book.
+URL: https://github.com/snlab-ch/migraph
+BugReports: https://github.com/snlab-ch/migraph/issues
 Depends: R (>= 4.0.0)
 License: MIT + file LICENSE
+Language: en-GB
 Encoding: UTF-8
 LazyData: true
 RoxygenNote: 7.1.1
@@ -22,7 +31,8 @@ Imports:
     dplyr,
     purrr,
     tibble,
-    tidyr
+    tidyr,
+    network
 Suggests: 
     testthat,
     roxygen2,
@@ -32,8 +42,7 @@ Suggests:
 Authors@R: 
     c(person(given = "James",
              family = "Hollway",
-             role = c("cre", "aut", "ctb"),
+             role = c("cph", "cre", "aut", "ctb"),
              email = "james.hollway@graduateinstitute.ch",
              comment = c("IHEID", ORCID = "0000-0002-8361-9647")))
-VignetteBuilder: knitr
 Roxygen: list(markdown = TRUE)

NAMESPACE

@@ -9,8 +9,10 @@ export("%>%")
 export(.E)
 export(.G)
 export(.N)
+export(as.network)
 export(as_igraph)
 export(as_matrix)
+export(as_network)
 export(as_tidygraph)
 export(blockmodel_concor)
 export(create_complete)
@@ -22,6 +24,7 @@ export(graph_closeness)
 export(graph_clustering)
 export(graph_degree)
 export(is.igraph)
+export(is.network)
 export(is.tbl_graph)
 export(is_bipartite)
 export(is_twomode)
@@ -62,6 +65,10 @@ importFrom(igraph,is.igraph)
 importFrom(igraph,is_bipartite)
 importFrom(igraph,mean_distance)
 importFrom(magrittr,"%>%")
+importFrom(network,as.matrix.network.adjacency)
+importFrom(network,as.matrix.network.incidence)
+importFrom(network,as.network)
+importFrom(network,is.network)
 importFrom(purrr,map)
 importFrom(rlang,.data)
 importFrom(rlang,enquo)

NEWS.md

@@ -1,3 +1,17 @@
+# migraph 0.6.1
+
+2021-04-11
+
+## Package
+
+* Closed #21 by elaborating DESCRIPTION file in preparation for CRAN submission
+* Updated several old URLs in documentation 
+
+## Classes
+
+* Closed #85 by adding `as_network()` to coerce objects into network class
+* Modified other coercion functions to also work with network class objects
+
 # migraph 0.6.0
 
 2021-03-03

R/convert.R

@@ -49,6 +49,12 @@ as_matrix <- function(object){
     } else {
       mat <- igraph::as_adjacency_matrix(object, sparse = FALSE)
     }
+  } else if (is.network(object)) {
+    if(network::is.bipartite(object)){
+      mat <- network::as.matrix.network.incidence(object)
+    } else {
+      mat <- network::as.matrix.network.adjacency(object)
+    }
   } else if (is.matrix(object)) {
     mat <- object
   } else if (is.data.frame(object)){
@@ -75,6 +81,7 @@ as_matrix <- function(object){
 }
 
 #' @rdname convert
+#' @importFrom network as.matrix.network.incidence as.matrix.network.adjacency
 #' @return An igraph graph object.
 #' @examples
 #' test <- data.frame(id1 = c("A","B","B","C","C"),
@@ -90,6 +97,14 @@ as_igraph <- function(object, twomode = FALSE){
     weights <- rlang::eval_tidy(weights, .E())
   } else if (is.igraph(object)) {
     graph <- object
+  } else if (is.network(object)) {
+    if(network::is.bipartite(object)){
+      graph <- network::as.matrix.network.incidence(object)
+      graph <- igraph::graph_from_incidence_matrix(graph)
+    } else {
+      graph <- network::as.matrix.network.adjacency(object)
+      graph <- igraph::graph_from_adjacency_matrix(graph)
+    }
   } else if (is.data.frame(object) | is.matrix(object)) {
     if (is.data.frame(object)) object <- as_matrix(object)
     if(nrow(object)!=ncol(object) | twomode){
@@ -115,6 +130,16 @@ as_tidygraph <- function(object, twomode = FALSE){
     tidy <- object
   } else if (is.igraph(object)) {
     tidy <- tidygraph::as_tbl_graph(object)
+  } else if (is.network(object)) {
+    if(network::is.bipartite(object)){
+      tidy <- network::as.matrix.network.incidence(object)
+      tidy <- igraph::graph_from_incidence_matrix(tidy)
+      tidy <- tidygraph::as_tbl_graph(tidy)
+    } else {
+      tidy <- network::as.matrix.network.adjacency(object)
+      tidy <- igraph::graph_from_adjacency_matrix(tidy)
+      tidy <- tidygraph::as_tbl_graph(tidy)
+    }
   } else if (is.data.frame(object) | is.matrix(object)) {
     if (is.data.frame(object)) object <- as_matrix(object)
     if(nrow(object)!=ncol(object) | twomode){
@@ -126,6 +151,31 @@ as_tidygraph <- function(object, twomode = FALSE){
   tidy
 }
 
+#' @rdname convert
+#' @importFrom network is.network as.network
+#' @return A sna/network network class object
+#' @examples
+#' test <- data.frame(id1 = c("A","B","B","C","C"),
+#'                    id2 = c("I","G","I","G","H"))
+#' as_network(test)
+#' @export
+as_network <- function(object){
+  
+  if(!network::is.network(object)){
+    if(is_twomode(object)){
+      net <- as_matrix(object)
+      net <- network::as.network(net, bipartite = TRUE)
+    } else {
+      net <- as_matrix(object)
+      net <- network::as.network(net, bipartite = FALSE)
+    }
+  } else {
+    net <- object
+  }
+  net
+  
+}
+
 #' @rdname convert
 #' @importFrom igraph is.bipartite
 #' @return A logical vector whether object is two-mode (TRUE) or not (FALSE)

R/data_elite_usa_advice.R

@@ -13,7 +13,7 @@
 #' @usage data(mpn_elite_usa_advice)
 #' @format Matrix with 14 rows and 20 columns
 #' @references
-#' Domhoff, G William. 2016. \href{http://www2.ucsc.edu/whorulesamerica/power_elite/}{“Who Rules America? Power Elite Database.”}
+#' Domhoff, G William. 2016. \href{https://whorulesamerica.ucsc.edu/power_elite/}{“Who Rules America? Power Elite Database.”}
 #' 
 #' Knoke, David, Mario Diani, James Hollway, and Dimitris C Christopoulos. 2021. 
 #' \href{https://www.cambridge.org/core/books/multimodal-political-networks/43EE8C192A1B0DCD65B4D9B9A7842128}{\emph{Multimodal Political Networks}}. 

R/data_elite_usa_money.R

@@ -13,7 +13,7 @@
 #' @usage data(mpn_elite_usa_money)
 #' @format Matrix with 26 rows and 6+6 columns
 #' @references 
-#' Domhoff, G William. 2016. \href{http://www2.ucsc.edu/whorulesamerica/power_elite/}{“Who Rules America? Power Elite Database.”}
+#' Domhoff, G William. 2016. \href{https://whorulesamerica.ucsc.edu/power_elite/}{“Who Rules America? Power Elite Database.”}
 #' 
 #' The Center for Responsive Politics. 2019. \href{http://www.opensecrets.org}{“OpenSecrets.”}
 #' 

R/reexports_network.R

@@ -0,0 +1,7 @@
+#' @importFrom network is.network
+#' @export
+network::is.network
+
+#' @importFrom network as.network
+#' @export
+network::as.network

README.md

@@ -23,18 +23,18 @@ Cambridge University Press: Cambridge.
 
 The package is offered as a complement to existing R packages for network analysis.
 It can analyse data in base formats such as matrices and (data frame) edgelists,
-but also leverages [`{igraph}`](https://igraph.org/r/) and is consistent with a [`{tidygraph}`](https://tidygraph.data-imaginist.com/index.html) workflow.
+but can also work with [`{igraph}`](https://igraph.org/r/) and [`{network}`](http://statnet.org) objects,
+and is consistent with a [`{tidygraph}`](https://tidygraph.data-imaginist.com/index.html) workflow.
 
-Please explore [the website](https://jhollway.github.io/migraph/) to find out more.
-Some of these functions are also relied on for the [Topological Typology](https://jhollway.shinyapps.io/TopoTypo/) app.
+Please explore [the website](https://snlab-ch.github.io/migraph/) to find out more.
 
 ## Installation
 
 ### From binary
 
 Perhaps the easiest way to install `{migraph}` is by installing a compiled binary.
 Binaries for all major OSes -- Windows, Mac, and Linux -- 
-can be found by clicking on the latest release [here](https://github.com/snlab-nl/rsiena/releases/latest).
+can be found by clicking on the latest release [here](https://github.com/snlab-ch/migraph/releases/latest).
 Download the appropriate binary for your operating system,
 and install using an adapted version of the following commands:
 
@@ -55,7 +55,7 @@ please install the `{remotes}` package from CRAN and then enter into the console
 It draws together, updates, and builds upon many functions currently available in
 other excellent R packages such as 
 [`{bipartite}`](https://github.com/biometry/bipartite), 
-[`{multinet}`](https://cran.r-project.org/web/packages/multinet/multinet.pdf), 
+[`{multinet}`](https://CRAN.R-project.org/package=multinet), 
 [`{netmem}`](https://github.com/anespinosa/netmem), 
 and [`{tnet}`](https://toreopsahl.com/tnet/),
 and implements many additional features currently only available outside the R ecosystem

cran-comments.md

@@ -0,0 +1,14 @@
+## Test environments
+
+* local R installation, R 4.0.3
+* Ubuntu Linux 20.04.1 LTS, R-release, GCC (on rhub)
+* Fedora Linux, R-devel, clang, gfortran (on rhub)
+* Mac OS X 10.15.7 (on Github), R 4.0.4
+* Microsoft Windows Server 2019 10.0.17763 (on Github), R 4.0.4
+* Ubuntu 20.04.2 (on Github), R 4.0.4
+
+## R CMD check results
+
+0 errors | 0 warnings | 0 notes
+
+* This is a new release.

tests/testthat/test-convert.R

@@ -27,3 +27,16 @@ test_that("as_matrix converts correctly",{
   expect_vector(mpn_elite_mex %>% as_matrix())
   expect_vector(mpn_elite_usa_advice %>% as_matrix())
 })
+
+test_that("as_tidygraph converts correctly",{
+  expect_s3_class(as_tidygraph(mat1), "tbl_graph")
+  expect_s3_class(as_tidygraph(southern_women), "tbl_graph")
+  expect_s3_class(as_tidygraph(mpn_elite_usa_money), "tbl_graph")
+})
+
+test_that("as_network converts correctly",{
+  expect_s3_class(as_network(mat1), "network")
+  expect_s3_class(as_network(southern_women), "network")
+  expect_s3_class(as_network(mpn_elite_usa_money), "network")
+})
+