day 16 python part 1 solve

this shouldn't work

day16/sol.py

@@ -13,37 +13,66 @@ for line in open("input"):
 
 print(G)
 
+def save():
+    seen = set()
+    with open("input.dot", "w") as f:
+        print("graph {", file=f)
+        for v, r, E in G:
+            if r:
+                print('%s [label="%s (%s)"]' % (v, v, r), file=f)
+            for e in E:
+                if (e,v) not in seen:
+                    print("%s -- %s" % (v,e), file=f)
+                    seen.add((v,e))
+        print("}", file=f)
+
+#save()
+
 #import astar
 
-score = {'AA': (0, frozenset())}
+score = {'AA': (0, 0, [])}
-
-#E = {x[0]: x[2] for x in G}
-#assert all(v in E[e] for v in E for e in E[v])
 
 # find the largest reward we can get in n steps
 # then find the largest reward we can get in n+1 steps
+potential = sum(r for _,r,_ in G)
 minutes = 30
-for _ in range(30):
+for _ in range(minutes):
     minutes -= 1
     next = {}
     for v, r, E in G:
-        s = set()
+        vo = v + 'o'
+        s = []
+        o = []
+        if vo in score:
+            reward, flow, open = score[vo]
+            o.append((reward, flow, open))
         if v in score:
-            reward, open = score[v]
+            reward, flow, open = score[v]
             # stay, don't open valve
-            s.add((reward, open))
+            s.append((reward, flow, open))
             # stay in place, open valve
             if v not in open:
                 reward += r*minutes
-                open = open.union(frozenset([v]))
+                o.append((reward, flow+r, open+[v]))
-                s.add((reward, open))
         # move here from somewhere else
         for e in E:
             if e in score:
-                s.add(score[e])
+                if v in score[e][-1]:
+                    o.append(score[e])
+                else:
+                    s.append(score[e])
+            eo = e+'o'
+            if eo in score:
+                if v in score[eo][-1]:
+                    o.append(score[eo])
+                else:
+                    s.append(score[eo])
         if s:
-            reward, open = max(s)
+            next[v] = max(s)
-            next[v] = (reward,open)
+        if o:
+            next[vo] = max(o)
     score = next
-    print(score)
+    print("%d minutes left" % minutes)
-print(max(r for r,_ in score.values()))
+    for v, (r, flow, open) in sorted(score.items(), key=lambda x: x[1]):
+        print("\t", v, r, "\t", ",".join(open))
+print(max(r for r,_,_ in score.values()))