Merge pull request #15 from OKN-CollabNext/skip-work-nodes

Don't display work nodes

collabnext/openalex/inference.py

@@ -0,0 +1,16 @@
+def infer_author_topic_edges(
+    author_work_edges: list[dict], work_topic_edges: list[dict]
+) -> list[dict]:
+    return [
+        {
+            "id": f"{author_work_edge['start']}-{work_topic_edge['end']}",
+            "start": author_work_edge["start"],
+            "end": work_topic_edge["end"],
+            "label": "TOPIC",
+            "start_type": "AUTHOR",
+            "end_type": "TOPIC",
+        }
+        for author_work_edge in author_work_edges
+        for work_topic_edge in work_topic_edges
+        if author_work_edge["end"] == work_topic_edge["start"]
+    ]

collabnext/openalex/institutions.py

@@ -3,4 +3,4 @@
 
 def get_institutions() -> list[Institution]:
     # Get 3 random institutions for now
-    return [Institutions().random() for _ in range(3)]
+    return [Institutions().random() for _ in range(5)]

observable/docs/data/graph.sqlite.py

@@ -10,6 +10,7 @@
     make_author_work_edges,
     make_work_topic_edges,
 )
+from collabnext.openalex.inference import infer_author_topic_edges
 from collabnext.openalex.institutions import (
     get_institutions,
 )
@@ -43,22 +44,29 @@
 # Create work nodes
 work_nodes = make_work_nodes(works)
 
-# Create work author edges
-work_author_edges = make_author_work_edges(works)
+# Create author-work edges
+author_work_edges = make_author_work_edges(works)
 
 # Get topics from works
 topics = get_work_topics(works)
 
 # Create topic nodes
 topic_nodes = make_topic_nodes(topics)
 
-# Create work topic edges
+# Create work-topic edges
 work_topic_edges = make_work_topic_edges(works)
 
+# Infer author-topic edges
+author_topic_edges = infer_author_topic_edges(author_work_edges, work_topic_edges)
 
 # Group all nodes and edges together
 nodes = [*institution_nodes, *author_nodes, *work_nodes, *topic_nodes]
-edges = [*affiliated_author_edges, *work_author_edges, *work_topic_edges]
+edges = [
+    *affiliated_author_edges,
+    *author_work_edges,
+    *author_topic_edges,
+    *work_topic_edges,
+]
 
 # Create nodes dataframe
 df_nodes = pd.DataFrame(nodes, columns=["id", "label", "type"])

observable/docs/index.md

@@ -26,6 +26,8 @@ const nodes = db.query(
     *
   FROM
     nodes
+  WHERE
+    nodes.type <> 'WORK'
   `
 );
 ```
@@ -37,6 +39,9 @@ const edges = db.query(
     *
   FROM
     edges
+  WHERE
+    edges.start_type <> 'WORK'
+    OR edges.end_type <> 'WORK'
   `
 );
 ```

tasks.py

@@ -54,4 +54,4 @@ def clean_branches(c):
 @task
 def touch(c):
     with cwd("observable/docs/data"):
-        c.run("touch graph.json.py")
+        c.run("touch graph.sqlite.py")