Changed LP so that minpay is only subtracted from terminal outcomes

Author: StephenPasteris

src/games/gameseq.cc

@@ -70,13 +70,13 @@ void GameSequenceForm::FillTableau(const GameNode &n, const Rational &prob,
                                    std::map<GamePlayer, GameSequence> &p_currentSequences)
 {
   if (n->GetOutcome()) {
-    IsOutcome(p_currentSequences) = 1;
     for (auto player : m_support.GetGame()->GetPlayers()) {
       GetPayoffEntry(p_currentSequences, player) +=
           prob * n->GetOutcome()->GetPayoff<Rational>(player);
     }
   }
   if (!n->GetInfoset()) {
+    IsOutcome(p_currentSequences) = 1;
     return;
   }
   if (n->GetPlayer()->IsChance()) {

src/solvers/lp/lp.cc

@@ -90,6 +90,9 @@ template <class T> void GameData<T>::FillTableau(Matrix<T> &A, const Game &p_gam
       if (p_game->IsOutcome(profile)) {
         A(row, col) = p_game->GetPayoff(profile, player1) - minpay;
       }
+      else {
+        A(row, col) = p_game->GetPayoff(profile, player1);
+      }
     }
   }
 }

test_internal_outcomes.py

@@ -0,0 +1,47 @@
+import pygambit as gbt
+
+solver = gbt.nash.lp_solve
+
+g = gbt.Game.new_tree(players=["1", "2"], title="2 player perfect info win lose")
+g.append_move(g.root, "2", ["a", "b"])
+g.append_move(g.root.children[0], "1", ["L", "R"])
+g.append_move(g.root.children[1], "1", ["L", "R"])
+g.append_move(g.root.children[0].children[0], "2", ["l", "r"])
+g.set_outcome(
+    g.root.children[0].children[0].children[0], g.add_outcome([1, -1], label="aLl")
+)
+g.set_outcome(
+    g.root.children[0].children[0].children[1], g.add_outcome([-1, 1], label="aLr")
+)
+g.set_outcome(g.root.children[0].children[1], g.add_outcome([1, -1], label="aR"))
+g.set_outcome(g.root.children[1].children[0], g.add_outcome([1, -1], label="bL"))
+g.set_outcome(g.root.children[1].children[1], g.add_outcome([-1, 1], label="bR"))
+
+print("blah")
+
+equilibria = solver(g).equilibria
+eqm = equilibria[0]
+print(eqm["1"])
+print(eqm["2"])
+
+
+g = gbt.Game.new_tree(players=["1", "2"], title="2 player perfect info win lose")
+g.append_move(g.root, "2", ["a", "b"])
+g.append_move(g.root.children[0], "1", ["L", "R"])
+g.append_move(g.root.children[1], "1", ["L", "R"])
+g.append_move(g.root.children[0].children[0], "2", ["l", "r"])
+g.set_outcome(g.root.children[0], g.add_outcome([-10, 10], label="a"))
+g.set_outcome(
+    g.root.children[0].children[0].children[0], g.add_outcome([11, -11], label="aLl")
+)
+g.set_outcome(
+    g.root.children[0].children[0].children[1], g.add_outcome([9, -9], label="aLr")
+)
+g.set_outcome(g.root.children[0].children[1], g.add_outcome([11, -11], label="aR"))
+g.set_outcome(g.root.children[1].children[0], g.add_outcome([1, -1], label="bL"))
+g.set_outcome(g.root.children[1].children[1], g.add_outcome([-1, 1], label="bR"))
+
+equilibria = solver(g).equilibria
+eqm = equilibria[0]
+print(eqm["1"])
+print(eqm["2"])