[RTM when tests pass] Refactor solution with classmethods

bioptim/examples/getting_started/example_simulation.py

@@ -27,7 +27,7 @@ def main():
     X = InitialGuess([0, 0, 0, 0])
     U = InitialGuess([-1, 1])
 
-    sol_from_initial_guess = Solution(ocp, [X, U])
+    sol_from_initial_guess = Solution.from_initial_guess(ocp, [X, U])
     s = sol_from_initial_guess.integrate(shooting_type=Shooting.SINGLE, integrator=SolutionIntegrator.OCP)
     print(f"Final position of q from single shooting of initial guess = {s.states['q'][:, -1]}")
     # Uncomment the next line to animate the integration
@@ -37,7 +37,7 @@ def main():
     U = np.random.rand(2, 31)
     U = InitialGuess(U, interpolation=InterpolationType.EACH_FRAME)
 
-    sol_from_initial_guess = Solution(ocp, [X, U])
+    sol_from_initial_guess = Solution.from_initial_guess(ocp, [X, U])
     s = sol_from_initial_guess.integrate(shooting_type=Shooting.SINGLE, integrator=SolutionIntegrator.OCP)
     print(f"Final position of q from single shooting of initial guess = {s.states['q'][:, -1]}")
     # Uncomment the next line to animate the integration

bioptim/examples/torque_driven_ocp/example_soft_contact.py

@@ -78,7 +78,7 @@ def initial_states_from_single_shooting(model, ns, tf, ode_solver):
     p = InitialGuessList()
     s = InitialGuessList()
 
-    sol_from_initial_guess = Solution(ocp, [x, u, p, s])
+    sol_from_initial_guess = Solution.from_initial_guess(ocp, [x, u, p, s])
     s = sol_from_initial_guess.integrate(shooting_type=Shooting.SINGLE, integrator=SolutionIntegrator.OCP)
     # s.animate()
 
@@ -92,7 +92,7 @@ def initial_states_from_single_shooting(model, ns, tf, ode_solver):
     p = InitialGuessList()
     s = InitialGuessList()
 
-    sol_from_initial_guess = Solution(ocp, [x, u, p, s])
+    sol_from_initial_guess = Solution.from_initial_guess(ocp, [x, u, p, s])
     s = sol_from_initial_guess.integrate(shooting_type=Shooting.SINGLE, integrator=SolutionIntegrator.OCP)
     # s.animate()
 

bioptim/gui/plot.py

@@ -687,7 +687,7 @@ def update_data(self, v: dict):
 
         self.ydata = []
 
-        sol = Solution(self.ocp, v)
+        sol = Solution.from_vector(self.ocp, v)
 
         if all([nlp.ode_solver.is_direct_collocation for nlp in self.ocp.nlp]):
             # no need to integrate when using direct collocation

bioptim/optimization/non_linear_program.py

@@ -2,11 +2,12 @@
 
 import casadi
 from casadi import SX, MX, Function, horzcat
+import numpy as np
 
 from .optimization_variable import OptimizationVariable, OptimizationVariableContainer
 from ..dynamics.ode_solver import OdeSolver
 from ..limits.path_conditions import InitialGuessList, BoundsList
-from ..misc.enums import ControlType, PhaseDynamics
+from ..misc.enums import ControlType, PhaseDynamics, Shooting
 from ..misc.options import OptionList
 from ..misc.mapping import NodeMapping
 from ..dynamics.dynamics_evaluation import DynamicsEvaluation

bioptim/optimization/optimal_control_program.py

@@ -1584,7 +1584,7 @@ def solve(
         self.ocp_solver.solve()
         self.is_warm_starting = False
 
-        return Solution(self, self.ocp_solver.get_optimized_value())
+        return Solution.from_dict(self, self.ocp_solver.get_optimized_value())
 
     def set_warm_start(self, sol: Solution):
         """

bioptim/optimization/receding_horizon_optimization.py

@@ -236,7 +236,7 @@ def _initialize_solution(self, states: list, controls: list):
         )
         s_init = InitialGuessList()
         p_init = InitialGuessList()
-        return Solution(solution_ocp, [x_init, u_init_for_solution, p_init, s_init])
+        return Solution.from_initial_guess(solution_ocp, [x_init, u_init_for_solution, p_init, s_init])
 
     def advance_window(self, sol: Solution, steps: int = 0, **advance_options):
         state_bounds_have_changed = self.advance_window_bounds_states(sol, **advance_options)
@@ -456,7 +456,7 @@ def _initialize_solution(self, states: list, controls: list):
         )
         s_init = InitialGuessList()
         p_init = InitialGuessList()
-        return Solution(solution_ocp, [x_init, u_init_for_solution, p_init, s_init])
+        return Solution.from_initial_guess(solution_ocp, [x_init, u_init_for_solution, p_init, s_init])
 
     def _initialize_state_idx_to_cycle(self, options):
         if "states" not in options:
@@ -691,7 +691,7 @@ def _initialize_solution(self, states: list, controls: list):
         )
         s_init = InitialGuessList()
         p_init = InitialGuessList()
-        return Solution(solution_ocp, [x_init, u_init_for_solution, p_init, s_init])
+        return Solution.from_initial_guess(solution_ocp, [x_init, u_init_for_solution, p_init, s_init])
 
     def _initialize_one_cycle(self, states: np.ndarray, controls: np.ndarray):
         """return a solution for a single window kept of the MHE"""
@@ -731,7 +731,7 @@ def _initialize_one_cycle(self, states: np.ndarray, controls: np.ndarray):
         )
         s_init = InitialGuessList()
         p_init = InitialGuessList()
-        return Solution(solution_ocp, [x_init, u_init_for_solution, p_init, s_init])
+        return Solution.from_initial_guess(solution_ocp, [x_init, u_init_for_solution, p_init, s_init])
 
 
 class NonlinearModelPredictiveControl(RecedingHorizonOptimization):