Remove the empty artifact from all types of networks

Augmenting the changes from PR se-sic#149, the empty artifact, which results from untracked files, is not a part of any network type anymore. This is change is based on commit d9f527c and fixes se-sic#153. This is a preliminary implementation with regard to issue se-sic#154. Signed-off-by: Claus Hunsen <[email protected]>
fehnkera · Sep 23, 2020 · 034bc76 · 034bc76
1 parent 68c4715
commit 034bc76
Show file tree

Hide file tree

Showing 4 changed files with 190 additions and 36 deletions.
diff --git a/tests/test-networks-artifact.R b/tests/test-networks-artifact.R
@@ -35,25 +35,13 @@ if (!dir.exists(CF.DATA)) CF.DATA = file.path(".", "tests", "codeface-data")
 
 test_that("Network construction of the undirected artifact-cochange network", {
 
-    ## configurations
-    proj.conf = ProjectConf$new(CF.DATA, CF.SELECTION.PROCESS, CASESTUDY, ARTIFACT)
-    proj.conf$update.value("commits.filter.base.artifact", FALSE)
-    net.conf = NetworkConf$new()
-    net.conf$update.values(updated.values = list(artifact.relation = "cochange"))
-
-    ## construct objects
-    proj.data = ProjectData$new(project.conf = proj.conf)
-    network.builder = NetworkBuilder$new(project.data = proj.data, network.conf = net.conf)
-
-    ## build network
-    network.built = network.builder$get.artifact.network()
-
-    ## vertex attributes
+    ## build expected network:
+    ## 1) vertices
     vertices = data.frame(name = c("Base_Feature", "foo", "A"),
                           kind = "Feature",
                           type = TYPE.ARTIFACT)
-
-    data = data.frame(
+    ## 2) edges
+    edges = data.frame(
         from = c("Base_Feature", "Base_Feature"),
         to = c("foo", "foo"),
         date = get.date.from.string(c("2016-07-12 16:06:32", "2016-07-12 16:06:32")),
@@ -65,9 +53,50 @@ test_that("Network construction of the undirected artifact-cochange network", {
         type = TYPE.EDGES.INTRA,
         relation = "cochange"
     )
+    ## 3) build expected network
+    network.expected = igraph::graph.data.frame(edges, directed = FALSE, vertices = vertices)
+
 
-    ## build expected network
-    network.expected = igraph::graph.data.frame(data, directed = FALSE, vertices = vertices)
+    ##
+    ## without untracked files
+    ##
+
+    ## configurations
+    proj.conf = ProjectConf$new(CF.DATA, CF.SELECTION.PROCESS, CASESTUDY, ARTIFACT)
+    proj.conf$update.value("commits.filter.base.artifact", FALSE)
+    proj.conf$update.value("commits.filter.untracked.files", TRUE)
+    net.conf = NetworkConf$new()
+    net.conf$update.values(updated.values = list(artifact.relation = "cochange"))
+
+    ## construct objects
+    proj.data = ProjectData$new(project.conf = proj.conf)
+    network.builder = NetworkBuilder$new(project.data = proj.data, network.conf = net.conf)
+
+    ## build network
+    network.built = network.builder$get.artifact.network()
+
+    ## test
+    expect_true(igraph::identical_graphs(network.built, network.expected))
+
+
+    ##
+    ## with untracked files
+    ##
+
+    ## configurations
+    proj.conf = ProjectConf$new(CF.DATA, CF.SELECTION.PROCESS, CASESTUDY, ARTIFACT)
+    proj.conf$update.value("commits.filter.base.artifact", FALSE)
+    proj.conf$update.value("commits.filter.untracked.files", FALSE)
+    net.conf = NetworkConf$new()
+    net.conf$update.values(updated.values = list(artifact.relation = "cochange"))
+
+    ## construct objects
+    proj.data = ProjectData$new(project.conf = proj.conf)
+    network.builder = NetworkBuilder$new(project.data = proj.data, network.conf = net.conf)
+
+    ## build network
+    network.built = network.builder$get.artifact.network()
 
+    ## test
     expect_true(igraph::identical_graphs(network.built, network.expected))
 })
diff --git a/tests/test-networks-author.R b/tests/test-networks-author.R
@@ -142,6 +142,7 @@ test_that("Amount of authors (author.all.authors, author.only.committers).", {
                 ## configurations
                 proj.conf = ProjectConf$new(CF.DATA, CF.SELECTION.PROCESS, CASESTUDY, ARTIFACT)
                 proj.conf$update.value("commits.filter.base.artifact", FALSE)
+                proj.conf$update.value("commits.filter.untracked.files", TRUE)
                 net.conf = NetworkConf$new()
 
                 ## update network configuration
@@ -198,25 +199,13 @@ test_that("Amount of authors (author.all.authors, author.only.committers).", {
 
 test_that("Network construction of the undirected author-cochange network", {
 
-    ## configurations
-    proj.conf = ProjectConf$new(CF.DATA, CF.SELECTION.PROCESS, CASESTUDY, ARTIFACT)
-    proj.conf$update.value("commits.filter.base.artifact", FALSE)
-    net.conf = NetworkConf$new()
-    net.conf$update.values(updated.values = list(author.relation = "cochange"))
-
-    ## construct objects
-    proj.data = ProjectData$new(project.conf = proj.conf)
-    network.builder = NetworkBuilder$new(project.data = proj.data, network.conf = net.conf)
-
-    ## build network
-    network.built = network.builder$get.author.network()
-
-    ## vertex attributes
+    ## build expected network:
+    ## 1) vertices
     authors = data.frame(name = c("Björn", "Olaf", "Karl", "Thomas"),
                          kind = TYPE.AUTHOR,
                          type = TYPE.AUTHOR)
 
-    ## edge attributes
+    ## 2) edges
     data = data.frame(comb.1. = c("Björn", "Björn", "Olaf", "Olaf", "Olaf", "Olaf", "Karl", "Karl"),
                       comb.2. = c("Olaf", "Olaf", "Karl", "Karl", "Thomas", "Thomas", "Thomas", "Thomas"),
                       date = get.date.from.string(c("2016-07-12 15:58:59", "2016-07-12 16:00:45", "2016-07-12 16:05:41",
@@ -234,10 +223,51 @@ test_that("Network construction of the undirected author-cochange network", {
                       type = TYPE.EDGES.INTRA,
                       relation = "cochange"
     )
-
-    ## build expected network
+    ## 3) build expected network
     network.expected = igraph::graph.data.frame(data, directed = FALSE, vertices = authors)
 
+
+    ##
+    ## without untracked files
+    ##
+
+    ## configurations
+    proj.conf = ProjectConf$new(CF.DATA, CF.SELECTION.PROCESS, CASESTUDY, ARTIFACT)
+    proj.conf$update.value("commits.filter.base.artifact", FALSE)
+    proj.conf$update.value("commits.filter.untracked.files", FALSE)
+    net.conf = NetworkConf$new()
+    net.conf$update.values(updated.values = list(author.relation = "cochange"))
+
+    ## construct objects
+    proj.data = ProjectData$new(project.conf = proj.conf)
+    network.builder = NetworkBuilder$new(project.data = proj.data, network.conf = net.conf)
+
+    ## build network
+    network.built = network.builder$get.author.network()
+
+    ## test
+    expect_true(igraph::identical_graphs(network.built, network.expected))
+
+
+    ##
+    ## with untracked files
+    ##
+
+    ## configurations
+    proj.conf = ProjectConf$new(CF.DATA, CF.SELECTION.PROCESS, CASESTUDY, ARTIFACT)
+    proj.conf$update.value("commits.filter.base.artifact", FALSE)
+    proj.conf$update.value("commits.filter.untracked.files", TRUE)
+    net.conf = NetworkConf$new()
+    net.conf$update.values(updated.values = list(author.relation = "cochange"))
+
+    ## construct objects
+    proj.data = ProjectData$new(project.conf = proj.conf)
+    network.builder = NetworkBuilder$new(project.data = proj.data, network.conf = net.conf)
+
+    ## build network
+    network.built = network.builder$get.author.network()
+
+    ## test
     expect_true(igraph::identical_graphs(network.built, network.expected))
 })
 
@@ -557,3 +587,36 @@ test_that("Network construction of the undirected author-issue network with just
 
     expect_true(igraph::identical_graphs(network.built, network.expected))
 })
+
+
+test_that("Network construction with only untracked files (no edges expected)", {
+
+    ## configurations
+    proj.conf = ProjectConf$new(CF.DATA, CF.SELECTION.PROCESS, CASESTUDY, ARTIFACT)
+    proj.conf$update.value("commits.filter.base.artifact", FALSE)
+    proj.conf$update.value("commits.filter.untracked.files", FALSE)
+    net.conf = NetworkConf$new()
+    net.conf$update.values(updated.values = list(author.relation = "cochange"))
+    net.conf$clear.edge.attributes()
+
+    ## construct objects and keep just commits on untracked files
+    proj.data = ProjectData$new(project.conf = proj.conf)
+    commit.data = subset(proj.data$get.commits(), artifact == UNTRACKED.FILE.EMPTY.ARTIFACT)
+    proj.data$set.commits(commit.data)
+
+    ## build network
+    network.builder = NetworkBuilder$new(project.data = proj.data, network.conf = net.conf)
+    network.built = network.builder$get.author.network()
+
+    ## build expected network (two vertices, no edges)
+    vertices = data.frame(name = c("Karl", "Thomas"), kind = TYPE.AUTHOR, type = TYPE.AUTHOR)
+    edges = create.empty.edge.list()
+    network.expected = igraph::graph.data.frame(edges, directed = FALSE, vertices = vertices)
+    ## add missing edge attributes
+    network.expected = igraph::set.edge.attribute(network.expected, "weight", value = as.numeric(NA))
+    network.expected = igraph::set.edge.attribute(network.expected, "type", value = as.character(NA))
+    network.expected = igraph::set.edge.attribute(network.expected, "relation", value = as.character(NA))
+
+    ## test
+    expect_true(igraph::identical_graphs(network.built, network.expected))
+})
diff --git a/tests/test-networks-bipartite.R b/tests/test-networks-bipartite.R
@@ -512,3 +512,47 @@ test_that("Construction of the directed bipartite network for the featureexpress
 
               expect_true(igraph::identical_graphs(network.built, network.expected))
 })
+
+
+test_that("Network construction with only untracked files (no edges and artifacts expected)", {
+
+    ## configurations
+    proj.conf = ProjectConf$new(CF.DATA, CF.SELECTION.PROCESS, CASESTUDY, ARTIFACT)
+    proj.conf$update.value("commits.filter.base.artifact", FALSE)
+    proj.conf$update.value("commits.filter.untracked.files", FALSE)
+    net.conf = NetworkConf$new()
+    net.conf$update.values(updated.values = list(author.relation = "cochange"))
+    net.conf$clear.edge.attributes()
+
+    ## construct objects and keep just commits on untracked files
+    proj.data = ProjectData$new(project.conf = proj.conf)
+    commit.data = subset(proj.data$get.commits(), artifact == UNTRACKED.FILE.EMPTY.ARTIFACT)
+    proj.data$set.commits(commit.data)
+
+    ## build network
+    network.builder = NetworkBuilder$new(project.data = proj.data, network.conf = net.conf)
+    network.built = network.builder$get.bipartite.network()
+
+    ## build expected network (two vertices, no edges):
+    ## 1) construct vertices
+    vertices = data.frame(name = c("Karl", "Thomas"), kind = TYPE.AUTHOR, type = TYPE.AUTHOR)
+    ## 2) add edge first to obtain edge attributes
+    edges = network.expected.data = data.frame(
+                          from = c("Karl"),
+                          to   = c("Thomas"),
+                          date = get.date.from.unix.timestamp(0),
+                          artifact.type = "",
+                          weight = 0,
+                          type = TYPE.EDGES.INTRA,
+                          relation = "cochange"
+                      )
+    ## 3) construct network
+    network.expected = construct.network.from.edge.list(vertices = vertices, edge.list = edges, network.conf = net.conf,
+                                                        directed = net.conf$get.value("author.directed"))
+    ## 4) remove edge again
+    network.expected = igraph::delete.edges(network.expected, 1)
+
+    ## test
+    expect_true(igraph::identical_graphs(network.built, network.expected))
+})
+
diff --git a/util-networks.R b/util-networks.R
@@ -304,6 +304,13 @@ NetworkBuilder = R6::R6Class("NetworkBuilder",
                 directed = FALSE
             )
 
+            ## remove the artifact vertices stemming from untracked files if existing
+            if ("name" %in% igraph::list.vertex.attributes(artifacts.net) &&
+                length(igraph::V(artifacts.net)[name == UNTRACKED.FILE.EMPTY.ARTIFACT]) > 0) {
+
+                artifacts.net = igraph::delete.vertices(artifacts.net, UNTRACKED.FILE.EMPTY.ARTIFACT)
+            }
+
             ## store network
             private$artifacts.network.cochange = artifacts.net
             logging::logdebug("get.artifact.network.cochange: finished.")
@@ -728,8 +735,14 @@ NetworkBuilder = R6::R6Class("NetworkBuilder",
                 return(vertices)
             })
 
-            ## Merge network data and construct a vertex-only network for now
+            ## Merge network data and construct a vertex-only network for now:
+            ## 1) construct vertex data (without edges)
             vertex.data = merge.network.data(vertex.data = vertex.data, edge.data = NULL)[["vertices"]]
+            ## 2) remove empty artifact, if names are available
+            if ("name" %in% colnames(vertex.data)) {
+                vertex.data = subset(vertex.data, !(name == UNTRACKED.FILE.EMPTY.ARTIFACT & type == TYPE.ARTIFACT))
+            }
+            ## 3) construct network without edges
             vertex.network = igraph::graph.data.frame(
                 d = create.empty.edge.list(),
                 vertices = vertex.data,
@@ -1177,7 +1190,12 @@ add.edges.for.bipartite.relation = function(net, bipartite.relations, network.co
 
         ## construct edges (i.e., a vertex sequence with c(source, target, source, target, ...))
         vertex.sequence.for.edges = parallel::mcmapply(function(d, a.df) {
+            ## get list of artifacts from edge list
             a = a.df[["data.vertices"]]
+
+            ## remove empty artifact
+            a = a[a != UNTRACKED.FILE.EMPTY.ARTIFACT]
+
             new.edges = lapply(a, function(vert) {
                 igraph::V(net)[d, vert] # get two vertices from source network:  c(author, artifact)
             })