SciX ID support for networks

vis_services/lib/author_network.py

@@ -78,7 +78,7 @@ def augment_graph_data(author_graph, data):
         return {"fullGraph" : author_graph}
 
     # make two dicts: a dict with all info for a certain paper (bib_dict)
-    # and a dict for each author that lists all bibcodes they helped write
+    # and a dict for each author that lists all identifiers they helped write
     name_dict = defaultdict(list)
     bib_dict= {}
 
@@ -90,12 +90,12 @@ def augment_graph_data(author_graph, data):
         read_count = d.get("read_count")
         if not read_count:
             read_count = 0
-        bib_dict[d.get("bibcode")] = {"title" : d.get("title", [""])[0],
+        bib_dict[d.get("scix_id")] = {"title" : d.get("title", [""])[0],
                                         "citation_count": citation_count,
                                         "read_count": read_count,
                                         "authors" : d.get("author_norm", [])}
         for author in d.get("author_norm", []):
-            name_dict[author].append(d.get("bibcode"))
+            name_dict[author].append(d.get("scix_id"))
     #don't allow any more author nodes than max_num_auth_nodes
     if len(author_graph['nodes'])>max_num_auth_nodes:
         node_cutoff = sorted([d["nodeWeight"] for d in author_graph['nodes']], reverse=True)[:max_num_auth_nodes+1][-1]
@@ -147,8 +147,8 @@ def augment_graph_data(author_graph, data):
                 name = G.nodes[child].get("nodeName")
                 size = G.nodes[child].get("nodeWeight")
                 bibs = sorted(name_dict[name], key=lambda x:bib_dict[x]["citation_count"], reverse=True)
-                total_citations = sum([bib_dict[bibcode]["citation_count"] for bibcode in name_dict[name]])
-                total_reads = sum([bib_dict[bibcode]["read_count"] for bibcode in name_dict[name]])
+                total_citations = sum([bib_dict[scix_id]["citation_count"] for scix_id in name_dict[name]])
+                total_reads = sum([bib_dict[scix_id]["read_count"] for scix_id in name_dict[name]])
                 children.append({"name":name, "size": size, "papers":bibs, "citation_count" : total_citations, "read_count" : total_reads, "numberName" : child})
             else:
                 G.remove_node(child)

vis_services/lib/paper_network.py

@@ -27,12 +27,12 @@
 # Helper functions
 def _get_reference_mapping(data):
     '''
-    Construct the reference dictionary for a set of bibcodes
+    Construct the reference dictionary for a set of SciX IDs
     '''
     refdict = {}
     for doc in data:
         if 'reference' in doc:
-            refdict[doc['bibcode']] = set(doc['reference'])
+            refdict[doc['scix_id']] = set(doc['reference'])
     return refdict
 
 def _get_paper_data(data):
@@ -41,7 +41,7 @@ def _get_paper_data(data):
     '''
     infodict = {}
     for doc in data:
-        infodict[doc['bibcode']] = doc
+        infodict[doc['scix_id']] = doc
     return infodict
 
 def _sort_and_cut_results(resdict,cutoff=1500):
@@ -189,7 +189,7 @@ def augment_graph_data(data, max_groups):
 # Main machinery
 def get_papernetwork(solr_data, max_groups, weighted=True, equalization=False, do_cutoff=False):
     '''
-    Given a list of bibcodes, this function builds the papers network based on co-citations
+    Given a list of identifiers, this function builds the papers network based on co-citations
     If 'weighted' is true, we will normalize the co-occurence frequency with the total number
     of papers in the set, otherwise we will work with the actual co-occurence frequencies.
     If 'equalization' is true, histogram equalization will be applied to the force values in
@@ -205,7 +205,7 @@ def get_papernetwork(solr_data, max_groups, weighted=True, equalization=False, d
               of the number of references in the linked nodes.
     '''
     # Get get paper list from the Solr data
-    papers_list = [a['bibcode'] for a in solr_data]
+    papers_list = [a['scix_id'] for a in solr_data]
     # First construct the reference dictionary, and a unique list of cited papers
     reference_dictionary = _get_reference_mapping(solr_data)
     # From now on we'll only work with publications that actually have references
@@ -287,11 +287,11 @@ def get_papernetwork(solr_data, max_groups, weighted=True, equalization=False, d
     #because the nodes must be inserted at the proper index
     nodes = [None]*len(ref_papers)
     for paper in solr_data:
-        if paper['bibcode'] not in selected_papers:
+        if paper['scix_id'] not in selected_papers:
             continue
-        index = ref_papers[paper["bibcode"]]
-        nodes[index] ={'nodeName':paper['bibcode'],
-                        'nodeWeight':paper.get('citation_count',1),
+        index = ref_papers[paper["scix_id"]]
+        nodes[index] ={'nodeName':paper['scix_id'],
+                       'nodeWeight':paper.get('citation_count',1),
                         'citation_count':paper.get('citation_count',0),
                         'read_count':paper.get('read_count',0),
                         'title':paper.get('title','NA')[0],