fix(chain): flag unhandled binding collision #614

Author: lmeyerov

CHANGELOG.md

@@ -6,6 +6,11 @@ The changelog format is based on [Keep a Changelog](https://keepachangelog.com/e
 This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html) and all PyGraphistry-specific breaking changes are explictly noted here.
 
 ## [Development]
+
+### Fixes
+
+* Hop: Detect #614 of node id column name colliding with edge src/dst id column name and raise `NotImplementedError`
+
 ### Docs
 
 * Python remote mode notebook: Fixed engine results

graphistry/compute/hop.py

@@ -116,6 +116,11 @@ def hop(self: Plottable,
     g2 = self.materialize_nodes(engine=EngineAbstract(engine_concrete.value))
     logger.debug('materialized node/eddge types: %s, %s', type(g2._nodes), type(g2._edges))
 
+    if g2._node == g2._source:
+        raise NotImplementedError(f'Not supported: Node id column cannot currently have the same name as edge src column: {g2._node}')
+    if g2._node == g2._destination:
+        raise NotImplementedError(f'Not supported: Node id column cannot currently have the same name as edge dst column: {g2._node}')
+
     starting_nodes = nodes if nodes is not None else g2._nodes
 
     if g2._edge is None:

graphistry/tests/compute/test_hop.py

@@ -415,6 +415,28 @@ def test_hop_predicates_fail_destination_forward(self, g_long_forwards_chain: CG
         assert g2._edges[['s', 'd']].sort_values(['s', 'd']).to_dict(orient='records') == []
 
 
+def test_hop_binding_reuse():
+    edges_df = pd.DataFrame({'s': ['a', 'b'], 'd': ['b', 'c']})
+    nodes1_df = pd.DataFrame({'v': ['a', 'b', 'c']})
+    nodes2_df = pd.DataFrame({'s': ['a', 'b', 'c']})
+    nodes3_df = pd.DataFrame({'d': ['a', 'b', 'c']})
+    
+    g1 = CGFull().nodes(nodes1_df, 'v').edges(edges_df, 's', 'd')
+    g2 = CGFull().nodes(nodes2_df, 's').edges(edges_df, 's', 'd')
+    g3 = CGFull().nodes(nodes3_df, 'd').edges(edges_df, 's', 'd')
+
+    try:
+        g1_hop = g1.hop()
+        g2_hop = g2.hop()
+        g3_hop = g3.hop()
+    except NotImplementedError:
+        return
+
+    assert g1_hop._nodes.shape == g2_hop._nodes.shape
+    assert g1_hop._edges.shape == g2_hop._edges.shape    
+    assert g1_hop._nodes.shape == g3_hop._nodes.shape
+    assert g1_hop._edges.shape == g3_hop._edges.shape    
+
 def test_hop_simple_cudf_pd():
     nodes_df = pd.DataFrame({'id': [0, 1, 2], 'label': ['a', 'b', 'c']})
     edges_df = pd.DataFrame({'src': [0, 1, 2], 'dst': [1, 2, 0]})