Return the improvement in the NashConv function.

PiperOrigin-RevId: 278189241 Change-Id: I80c9e1f0ef729a48666fa6ad565afdb194b107e3
google-deepmind · Nov 4, 2019 · 222ba03 · 222ba03
1 parent c5884da
commit 222ba03
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/open_spiel/python/algorithms/exploitability.py b/open_spiel/python/algorithms/exploitability.py
@@ -190,4 +190,4 @@ def nash_conv(game, policy, return_only_nash_conv=True):
     return nash_conv_
   else:
     return _NashConvReturn(
-        nash_conv=nash_conv_, player_improvements=on_policy_values)
+        nash_conv=nash_conv_, player_improvements=player_improvements)