From b270cc1a44144d84faca5ae73b2d75d1a89c5f65 Mon Sep 17 00:00:00 2001
From: Zaccharie Ramzi <zaccharie.ramzi@gmail.com>
Date: Fri, 7 Oct 2022 20:10:20 +0200
Subject: [PATCH] added gamma to lbfgs state

---
 jaxopt/_src/lbfgs.py | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

diff --git a/jaxopt/_src/lbfgs.py b/jaxopt/_src/lbfgs.py
index 639a839a..9f79ada9 100644
--- a/jaxopt/_src/lbfgs.py
+++ b/jaxopt/_src/lbfgs.py
@@ -142,6 +142,7 @@ class LbfgsState(NamedTuple):
   s_history: Any
   y_history: Any
   rho_history: jnp.ndarray
+  gamma: jnp.ndarray
   aux: Optional[Any] = None
 
 
@@ -255,6 +256,7 @@ def init_state(self,
                       s_history=init_history(init_params, self.history_size),
                       y_history=init_history(init_params, self.history_size),
                       rho_history=jnp.zeros(self.history_size, dtype=dtype),
+                      gamma=jnp.asarray(1.0, dtype=dtype),
                       aux=aux)
 
   def update(self,
@@ -275,16 +277,10 @@ def update(self,
     (value, aux), grad = self._value_and_grad_with_aux(params, *args, **kwargs)
 
     start = state.iter_num % self.history_size
-    previous = (start + self.history_size - 1) % self.history_size
-
-    if self.use_gamma:
-      gamma = compute_gamma(state.s_history, state.y_history, previous)
-    else:
-      gamma = 1.0
 
     product = inv_hessian_product(pytree=grad, s_history=state.s_history,
                                   y_history=state.y_history,
-                                  rho_history=state.rho_history, gamma=gamma,
+                                  rho_history=state.rho_history, gamma=state.gamma,
                                   start=start)
     descent_direction = tree_scalar_mul(-1.0, product)
 
@@ -351,6 +347,11 @@ def update(self,
     y_history = update_history(state.y_history, y, last)
     rho_history = update_history(state.rho_history, rho, last)
 
+    if self.use_gamma:
+      gamma = compute_gamma(s_history, y_history, last)
+    else:
+      gamma = jnp.array(1.0)
+
     new_state = LbfgsState(iter_num=state.iter_num + 1,
                            value=new_value,
                            stepsize=jnp.asarray(new_stepsize),
@@ -358,6 +359,7 @@ def update(self,
                            s_history=s_history,
                            y_history=y_history,
                            rho_history=rho_history,
+                           gamma=gamma,
                            # FIXME: we should return new_aux here but
                            # BacktrackingLineSearch currently doesn't support
                            # an aux.