fix postprocessing

Author: chreman

.gitignore

@@ -34,4 +34,5 @@ coverage/
 /lc_browseview_cache.json
 /lc_cache.json
 /linkedcat.sqlite
-.env
+.env
+.Rproj.user

server/preprocessing/other-scripts/postprocess.R

@@ -9,12 +9,13 @@ create_overview_output <- function(named_clusters, layout, metadata) {
   cluster_labels = named_clusters$cluster_labels
 
   # Prepare the output
-  result = data.frame(cbind(x, y, labels, cluster_labels))
+  result = data.frame(cbind(x, y, labels, cluster_labels, metadata$id))
+  names(result)[5] <- "id"
   unique_groups = data.frame(unique(result$cluster_labels))
   colnames(unique_groups) <- "cluster_labels"
   unique_groups$groups <- seq_along(unique_groups$cluster_labels)
   result = merge(result, unique_groups, by='cluster_labels')
-  output = merge(metadata, result, by.x="id", by.y="labels", all=TRUE)
+  output = merge(metadata, result, by.x="id", by.y="id", all=TRUE)
 
   names(output)[names(output)=="groups"] <- "area_uri"
   output["area"] = paste(output$cluster_labels, sep="")

server/preprocessing/other-scripts/summarize.R

@@ -57,8 +57,7 @@ create_cluster_labels <- function(clusters, metadata,
   tfidf_top_names <- get_top_names(tfidf_top, top_n, stops)
   clusters$cluster_labels = ""
   for (k in seq(1, clusters$num_clusters)) {
-    group = c(names(clusters$groups[clusters$groups == k]))
-    matches = which(clusters$labels%in%group)
+    matches = which(unname(clusters$groups == k) == TRUE)
     summary = tfidf_top_names[[k]]
     if (summary == "") {
       candidates = mapply(paste, metadata$title[matches], metadata$paper_abstract[matches])
@@ -106,8 +105,7 @@ get_cluster_corpus <- function(clusters, metadata, service, stops, taxonomy_sepa
                                add_title_ngrams = T) {
   subjectlist = list()
   for (k in seq(1, clusters$num_clusters)) {
-    group = c(names(clusters$groups[clusters$groups == k]))
-    matches = which(metadata$id %in% group)
+    matches = which(unname(clusters$groups == k) == TRUE)
     titles =  metadata$title[matches]
     subjects = metadata$subject[matches]
     titles = lapply(titles, function(x) {gsub("[^[:alnum:]-]", " ", x)})