Generate a price-map for the optimizer based on the significance of t…

…he exchange which the price is taken from (B2, then B3, then U2, then U3, then all the rest with equal significance).
bancorprotocol · May 28, 2024 · 7b994d4 · 7b994d4
1 parent 7c58f59
commit 7b994d4
Show file tree

Hide file tree

Showing 3 changed files with 19 additions and 29 deletions.
diff --git a/fastlane_bot/modes/base.py b/fastlane_bot/modes/base.py
@@ -57,19 +57,28 @@ def get_profit(self, src_token: str, optimization, trade_instructions_df):
 
     def calculate_profit(self, src_token: str, src_profit: float) -> Decimal:
         if src_token not in [self.ConfigObj.NATIVE_GAS_TOKEN_ADDRESS, self.ConfigObj.WRAPPED_GAS_TOKEN_ADDRESS]:
-            price = get_reliable_price(self.CCm, self.sort_order, self.ConfigObj.WRAPPED_GAS_TOKEN_ADDRESS, src_token)
-            assert price is not None, f"Failed to get conversion rate for {src_token} and {self.ConfigObj.WRAPPED_GAS_TOKEN_ADDRESS}"
+            price = self.find_reliable_price(self.CCm, self.ConfigObj.WRAPPED_GAS_TOKEN_ADDRESS, src_token)
+            assert price is not None, f"No conversion rate for {self.ConfigObj.WRAPPED_GAS_TOKEN_ADDRESS} and {src_token}"
             return Decimal(str(src_profit)) / Decimal(str(price))
         return Decimal(str(src_profit))
 
+    def get_params(self, cc, dst_tokens, src_token):
+        pstart = {src_token: 1}
+        for dst_token in dst_tokens:
+            if dst_token != src_token:
+                pstart[dst_token] = self.find_reliable_price(cc, dst_token, src_token)
+                if pstart[dst_token] is None:
+                    return None
+        return {"pstart": pstart}
+
+    def find_reliable_price(self, cc, dst_token, src_token):
+        list1 = [{"exchange": curve.params.exchange, "price": curve.p / 1} for curve in cc.bytknx(dst_token).bytkny(src_token).curves]
+        list2 = [{"exchange": curve.params.exchange, "price": 1 / curve.p} for curve in cc.bytknx(src_token).bytkny(dst_token).curves]
+        items = sorted(list1 + list2, key = lambda item: self.sort_order.get(item["exchange"], float("inf")))
+        return items[0]["price"] if len(items) > 0 else None
+
 def is_net_change_small(trade_instructions_df) -> bool:
     try:
         return max(trade_instructions_df.iloc[-1]) < 1e-4
     except Exception:
         return False
-
-def get_reliable_price(CCm, sort_order, dst_token, src_token):
-    list1 = [{"exchange": curve.params.exchange, "price": curve.p / 1} for curve in CCm.bytknx(dst_token).bytkny(src_token).curves]
-    list2 = [{"exchange": curve.params.exchange, "price": 1 / curve.p} for curve in CCm.bytknx(src_token).bytkny(dst_token).curves]
-    items = sorted(list1 + list2, key = lambda item: sort_order.get(item["exchange"], float('inf')))
-    return items[0]["price"] if len(items) > 0 else None
diff --git a/fastlane_bot/modes/base_pairwise.py b/fastlane_bot/modes/base_pairwise.py
@@ -29,7 +29,7 @@ def find_arbitrage(self) -> Dict[List[Any], List[Any]]:
                 try:
                     container = CPCContainer(curve_combos)
                     optimizer = PairOptimizer(container)
-                    params = get_params(container, dst_token, src_token)
+                    params = self.get_params(container, [dst_token], src_token)
                     optimization = optimizer.optimize(src_token, params=params)
                     trade_instructions_dic = optimization.trade_instructions(optimizer.TIF_DICTS)
                     trade_instructions_df = optimization.trade_instructions(optimizer.TIF_DFAGGR)
@@ -45,7 +45,3 @@ def find_arbitrage(self) -> Dict[List[Any], List[Any]]:
                     arb_opps.append({"profit": profit, "src_token": src_token, "trade_instructions_dic": trade_instructions_dic})
 
         return {"combos": combos, "arb_opps": sorted(arb_opps, key=lambda arb_opp: arb_opp["profit"], reverse=True)}
-
-def get_params(container, dst_token, src_token):
-    pstart = {dst_token: container.bypairs(f"{dst_token}/{src_token}").curves[0].p}
-    return {"pstart": pstart}
diff --git a/fastlane_bot/modes/base_triangle.py b/fastlane_bot/modes/base_triangle.py
@@ -22,7 +22,7 @@ def find_arbitrage(self) -> Dict[List[Any], List[Any]]:
             try:
                 container = CPCContainer(miniverse)
                 optimizer = MargPOptimizer(container)
-                params = get_params(container, src_token)
+                params = self.get_params(container, container.tokens(), src_token)
                 optimization = optimizer.optimize(src_token, params=params)
                 trade_instructions_dic = optimization.trade_instructions(optimizer.TIF_DICTS)
                 trade_instructions_df = optimization.trade_instructions(optimizer.TIF_DFAGGR)
@@ -38,18 +38,3 @@ def find_arbitrage(self) -> Dict[List[Any], List[Any]]:
                 arb_opps.append({"profit": profit, "src_token": src_token, "trade_instructions_dic": trade_instructions_dic})
 
         return {"combos": combos, "arb_opps": sorted(arb_opps, key=lambda arb_opp: arb_opp["profit"], reverse=True)}
-
-def get_params(container, src_token):
-    pstart = {src_token: 1}
-    for dst_token in container.tokens():
-        if dst_token != src_token:
-            curves = container.bytknx(dst_token).bytkny(src_token).curves
-            if len(curves) > 0:
-                pstart[dst_token] = curves[0].p
-            else:
-                curves = container.bytknx(src_token).bytkny(dst_token).curves
-                if len(curves) > 0:
-                    pstart[dst_token] = 1 / curves[0].p
-                else:
-                    return None
-    return {"pstart": pstart}