set numeric_only=True for pandas 2.0 breaking changes

Author: a-r-j

graphein/protein/edges/distance.py

@@ -1210,7 +1210,7 @@ def get_ring_centroids(ring_atom_df: pd.DataFrame) -> pd.DataFrame:
     """
     return (
         ring_atom_df.groupby("node_id")
-        .mean()[["x_coord", "y_coord", "z_coord"]]
+        .mean(numeric_only=True)[["x_coord", "y_coord", "z_coord"]]
         .reset_index()
     )
 

graphein/protein/features/nodes/geometry.py

@@ -47,7 +47,7 @@ def add_sidechain_vector(
     if "rgroup_df" not in g.graph.keys():
         g.graph["rgroup_df"] = compute_rgroup_dataframe(g.graph["raw_pdb_df"])
 
-    sc_centroid = g.graph["rgroup_df"].groupby("node_id").mean()
+    sc_centroid = g.graph["rgroup_df"].groupby("node_id").mean(numeric_only=True)
 
     # Iterate over nodes and compute vector
     for n, d in g.nodes(data=True):

graphein/protein/graphs.py

@@ -1138,7 +1138,7 @@ def number_groups_of_runs(list_of_values: List[Any]) -> List[str]:
     """
     df = pd.DataFrame({"val": list_of_values})
     df["idx"] = df["val"].shift() != df["val"]
-    df["sum"] = df.groupby("val")["idx"].cumsum()
+    df["sum"] = df.groupby("val")["idx"].cumsum(numeric_only=True)
     return list(df["val"].astype(str) + df["sum"].astype(str))
 
 

graphein/protein/tensor/io.py

@@ -267,7 +267,7 @@ def protein_to_pyg(
 
     if store_bfactor:
         # group by residue_id and average b_factor per residue
-        residue_bfactors = df.groupby("residue_id")["b_factor"].mean()
+        residue_bfactors = df.groupby("residue_id")["b_factor"].mean(numeric_only=True)
         out.bfactor = torch.from_numpy(residue_bfactors.values)
 
     return out

tests/protein/nodes/features/test_geometry.py

@@ -138,7 +138,7 @@ def test_add_sidechain_vector():
             sc_true = np.array(
                 g.graph["rgroup_df"]
                 .groupby("node_id")
-                .mean()
+                .mean(numeric_only=True)
                 .loc[n][["x_coord", "y_coord", "z_coord"]]
             )
             np.testing.assert_almost_equal(