embedding data extraction

util/data_extraction.py

@@ -4,6 +4,9 @@ import dimod
 import random 
 import numpy as np
 
+from bson.objectid import ObjectId
+import pandas as pd
+
 from tqdm import tqdm
 
 def extract_wmis_qpu_results():
@@ -213,3 +216,73 @@ def extract_instance_parameters():
     edb.commit()
     edb_cursor.close()
     edb.close()
+
+def extract_embedding_data():
+    edb = script.connect_to_experimetns_db()
+    edb_cursor = edb.cursor()
+
+    idb = script.connect_to_instance_pool()
+
+    scope = input("scope: ")
+    table_name = input("table name: ")
+
+    qubo_collection = input("qubo collection: ")
+    solver_graph_id = ObjectId(input("solver graph id: "))
+
+    solver_input = queries.WMIS_solver_input_scope_query(idb, qubo_collection)
+    solver_input.query(scope, solver_graph_id)
+
+    results = pd.DataFrame(columns=["instance_id",
+                                    "embedding_list_id",
+                                    "emb_list_index",
+                                    "num_qubits",
+                                    "avgr_chain_length",
+                                    "median_chain_length"]);
+    
+    for soin in solver_input:
+        emb_index = 0;
+        for emb in soin["embeddings"]:
+            chain_lengths = []
+            qubits = set()
+
+            for node, chain in emb.items():
+                
+                chain_lengths.append(len(chain))
+
+                qubits = qubits.union(set(chain))
+
+            results = results.append({"instance_id": soin["instance_id"],
+                                      "embedding_list_id": soin["embeddings_id"],
+                                      "emb_list_index": emb_index,
+                                      "num_qubits": len(qubits),
+                                      "avgr_chain_length": np.mean(chain_lengths),
+                                      "median_chain_length": np.median(chain_lengths)},
+                                     ignore_index=True)
+
+    insert_row = ('''INSERT INTO {} 
+                     (instance_id, 
+                      embedding_list_id, 
+                      emb_list_index, 
+                      num_qubits, 
+                      avgr_chain_length, 
+                      median_chain_length) 
+                     VALUES(%s, %s, %s, %s, %s, %s)''').format(table_name)
+
+    for index, row in results.iterrows():
+        edb_cursor.execute(insert_row, (str(row["instance_id"]),
+                                        str(row["embedding_list_id"]),
+                                        int(row["emb_list_index"]),
+                                        int(row["num_qubits"]),
+                                        float(row["avgr_chain_length"]),
+                                        float(row["median_chain_length"])))
+    
+    edb.commit()
+    edb_cursor.close()
+    edb.close()
+
+
+
+
+        
+
+