Adapt to do 0.5.0

Full renaming of modules and classes in do 0.5.0 (to come).
airbus · Oct 18, 2024 · 88cce97 · 88cce97
1 parent a13d22d
commit 88cce97
Show file tree

Hide file tree

Showing 10 changed files with 1,941 additions and 1,970 deletions.
diff --git a/examples/scheduling/rcpsp_examples.py b/examples/scheduling/rcpsp_examples.py
@@ -1,4 +1,7 @@
-from discrete_optimization.rcpsp.rcpsp_solvers import CP_MRCPSP_MZN, CP_RCPSP_MZN
+from discrete_optimization.rcpsp.solvers_map import (
+    CpMultimodeRcpspSolver,
+    CpRcpspSolver,
+)
 
 from examples.scheduling.rcpsp_datasets import get_complete_path
 from examples.scheduling.rcpsp_multiskill_datasets import get_data_available_ms
@@ -33,11 +36,11 @@ def random_walk():
     print("end times: ")
     for task_id in states[-1].tasks_details.keys():
         print("end task", task_id, ": ", states[-1].tasks_details[task_id].end)
-    from discrete_optimization.rcpsp.rcpsp_utils import (
+    import matplotlib.pyplot as plt
+    from discrete_optimization.rcpsp.utils import (
         plot_resource_individual_gantt,
         plot_ressource_view,
         plot_task_gantt,
-        plt,
     )
 
     from skdecide.hub.solver.do_solver.sk_to_do_binding import (
@@ -63,7 +66,7 @@ def do_singlemode():
             delta_index_freedom=0,
             delta_time_freedom=0,
         ),
-        do_solver_type=CP_RCPSP_MZN,
+        do_solver_type=CpRcpspSolver,
     )
     solver.solve()
     states, actions, values = rollout(
@@ -92,7 +95,7 @@ def do_multimode():
             delta_index_freedom=0,
             delta_time_freedom=0,
         ),
-        do_solver_type=CP_MRCPSP_MZN,
+        do_solver_type=CpMultimodeRcpspSolver,
     )
     solver.solve()
     states, actions, values = rollout(

diff --git a/examples/scheduling/toy_rcpsp_examples.py b/examples/scheduling/toy_rcpsp_examples.py
@@ -475,11 +475,11 @@ def run_astar():
     )
 
     do_sol = from_last_state_to_solution(states[-1], domain)
-    from discrete_optimization.rcpsp.rcpsp_utils import (
+    import matplotlib.pyplot as plt
+    from discrete_optimization.rcpsp.utils import (
         plot_resource_individual_gantt,
         plot_ressource_view,
         plot_task_gantt,
-        plt,
     )
 
     plot_task_gantt(do_sol.problem, do_sol)
@@ -528,11 +528,11 @@ def run_do():
     )
 
     do_sol = from_last_state_to_solution(states[-1], domain)
-    from discrete_optimization.rcpsp.rcpsp_utils import (
+    import matplotlib.pyplot as plt
+    from discrete_optimization.rcpsp.utils import (
         plot_resource_individual_gantt,
         plot_ressource_view,
         plot_task_gantt,
-        plt,
     )
 
     plot_task_gantt(do_sol.problem, do_sol)

diff --git a/notebooks/13_scheduling_tuto.ipynb b/notebooks/13_scheduling_tuto.ipynb
@@ -869,7 +869,7 @@
     "from skdecide.hub.solver.do_solver.sk_to_do_binding import from_last_state_to_solution\n",
     "\n",
     "do_sol = from_last_state_to_solution(states[-1], domain)\n",
-    "from discrete_optimization.rcpsp.rcpsp_utils import plot_ressource_view, plot_task_gantt\n",
+    "from discrete_optimization.rcpsp.utils import plot_ressource_view, plot_task_gantt\n",
     "\n",
     "# Each line of the plot is a task\n",
     "fig_gantt = plot_task_gantt(do_sol.problem, do_sol)\n",

diff --git a/poetry.lock b/poetry.lock
diff --git a/pyproject.toml b/pyproject.toml
@@ -58,7 +58,7 @@ matplotlib = { version = ">=3.3.4", optional = true }
 joblib = { version = ">=1.0.1", optional = true }
 stable-baselines3 = { version = ">=2.0.0", optional = true }
 ray = { extras = ["rllib"], version = ">=2.9.0", optional = true }
-discrete-optimization = { version = ">=0.4.0" }
+discrete-optimization = { version = ">=0.5.0" }
 openap = { version = ">=1.5", python = ">=3.9", optional = true }
 pygeodesy = { version = ">=23.6.12", optional = true }
 unified-planning = { version = ">=1.1.0", python = ">=3.10", optional = true }

diff --git a/skdecide/hub/domain/rcpsp/rcpsp_sk_parser.py b/skdecide/hub/domain/rcpsp/rcpsp_sk_parser.py
@@ -13,12 +13,12 @@
 
 def load_domain(file_path):
     """"""
-    from discrete_optimization.rcpsp.rcpsp_model import RCPSPModel
-    from discrete_optimization.rcpsp.rcpsp_parser import parse_file
+    from discrete_optimization.rcpsp.parser import parse_file
+    from discrete_optimization.rcpsp.problem import RcpspProblem
 
     from skdecide.hub.domain.rcpsp.rcpsp_sk import MRCPSP, RCPSP
 
-    rcpsp_model: RCPSPModel = parse_file(file_path)
+    rcpsp_model: RcpspProblem = parse_file(file_path)
     if not rcpsp_model.is_rcpsp_multimode():
         my_domain = RCPSP(
             resource_names=rcpsp_model.resources_list,
@@ -50,9 +50,7 @@ def load_domain(file_path):
 
 
 def load_multiskill_domain(file_path):
-    from discrete_optimization.rcpsp_multiskill.rcpsp_multiskill_parser import (
-        parse_file,
-    )
+    from discrete_optimization.rcpsp_multiskill.parser import parse_file
 
     model_msrcpsp, new_tame_to_original_task_id = parse_file(
         file_path, max_horizon=2000

diff --git a/skdecide/hub/solver/do_solver/do_solver_scheduling.py b/skdecide/hub/solver/do_solver/do_solver_scheduling.py
@@ -19,11 +19,11 @@
 from discrete_optimization.generic_tools.result_storage.result_storage import (
     ResultStorage,
 )
-from discrete_optimization.rcpsp.rcpsp_model import RCPSPModel, RCPSPSolution
-from discrete_optimization.rcpsp_multiskill.rcpsp_multiskill import (
-    MS_RCPSPModel,
-    MS_RCPSPSolution,
-    MS_RCPSPSolution_Variant,
+from discrete_optimization.rcpsp.problem import RcpspProblem, RcpspSolution
+from discrete_optimization.rcpsp_multiskill.problem import (
+    MultiskillRcpspProblem,
+    MultiskillRcpspSolution,
+    VariantMultiskillRcpspSolution,
 )
 
 from skdecide import Domain
@@ -78,22 +78,21 @@ def build_solver(
     A class of do-solver, associated with some default parameters to be passed to its constructor and solve function
     (and potentially init_model function)
     """
-    if isinstance(do_domain, RCPSPModel):
-        from discrete_optimization.rcpsp.rcpsp_solvers import (
-            look_for_solver,
-            solvers_map,
-        )
+    if isinstance(do_domain, RcpspProblem):
+        from discrete_optimization.rcpsp.solvers_map import look_for_solver, solvers_map
 
-        do_domain_cls = RCPSPModel
-    elif isinstance(do_domain, MS_RCPSPModel):
-        from discrete_optimization.rcpsp_multiskill.rcpsp_multiskill_solvers import (
+        do_domain_cls = RcpspProblem
+    elif isinstance(do_domain, MultiskillRcpspProblem):
+        from discrete_optimization.rcpsp_multiskill.solvers_map import (
             look_for_solver,
             solvers_map,
         )
 
-        do_domain_cls = MS_RCPSPModel
+        do_domain_cls = MultiskillRcpspProblem
     else:
-        raise ValueError("do_domain should be either a RCPSPModel or a MS_RCPSPModel.")
+        raise ValueError(
+            "do_domain should be either a RcpspProblem or a MultiskillRcpspProblem."
+        )
     available = look_for_solver(do_domain)
     if solver_type is not None:
         if solver_type in solvers_map:
@@ -114,7 +113,9 @@ def build_solver(
 
 
 def from_solution_to_policy(
-    solution: Union[RCPSPSolution, MS_RCPSPSolution, MS_RCPSPSolution_Variant],
+    solution: Union[
+        RcpspSolution, MultiskillRcpspSolution, VariantMultiskillRcpspSolution
+    ],
     domain: SchedulingDomain,
     policy_method_params: PolicyMethodParams,
 ) -> PolicyRCPSP:
@@ -125,7 +126,7 @@ def from_solution_to_policy(
     schedule = None
     resource_allocation = None
     resource_allocation_priority = None
-    if isinstance(solution, RCPSPSolution):
+    if isinstance(solution, RcpspSolution):
         permutation_task = sorted(
             solution.rcpsp_schedule,
             key=lambda x: (solution.rcpsp_schedule[x]["start_time"], x),
@@ -137,7 +138,7 @@ def from_solution_to_policy(
         modes_dictionnary[solution.problem.n_jobs_non_dummy + 2] = 1
         for i in range(len(solution.rcpsp_modes)):
             modes_dictionnary[i + 2] = solution.rcpsp_modes[i]
-    elif isinstance(solution, MS_RCPSPSolution):
+    elif isinstance(solution, MultiskillRcpspSolution):
         permutation_task = sorted(
             solution.schedule, key=lambda x: (solution.schedule[x]["start_time"], x)
         )
@@ -149,7 +150,7 @@ def from_solution_to_policy(
             ]  # warning here...
             for task in solution.employee_usage
         }
-        if isinstance(solution, MS_RCPSPSolution_Variant):
+        if isinstance(solution, VariantMultiskillRcpspSolution):
             resource_allocation_priority = solution.priority_worker_per_task
             modes_dictionnary = {}
             # set modes for start and end (dummy) jobs
@@ -219,15 +220,15 @@ def __init__(
 
     def get_available_methods(self, domain: SchedulingDomain):
         do_domain = build_do_domain(domain)
-        if isinstance(do_domain, (MS_RCPSPModel)):
-            from discrete_optimization.rcpsp_multiskill.rcpsp_multiskill_solvers import (
+        if isinstance(do_domain, (MultiskillRcpspProblem)):
+            from discrete_optimization.rcpsp_multiskill.solvers_map import (
                 look_for_solver,
                 solvers_map,
             )
 
             available = look_for_solver(do_domain)
-        elif isinstance(do_domain, RCPSPModel):
-            from discrete_optimization.rcpsp.rcpsp_solvers import (
+        elif isinstance(do_domain, RcpspProblem):
+            from discrete_optimization.rcpsp.solvers_maps import (
                 look_for_solver,
                 solvers_map,
             )
@@ -263,7 +264,7 @@ def _solve(self) -> None:
             self.solver.init_model(**copy_dict_params)
 
         result_storage = self.solver.solve(callbacks=callbacks, **copy_dict_params)
-        best_solution: RCPSPSolution = result_storage.get_best_solution()
+        best_solution: RcpspSolution = result_storage.get_best_solution()
 
         assert best_solution is not None
 

diff --git a/skdecide/hub/solver/do_solver/gphh.py b/skdecide/hub/solver/do_solver/gphh.py
@@ -23,8 +23,8 @@
 from discrete_optimization.generic_tools.hyperparameters.hyperparametrizable import (
     Hyperparametrizable,
 )
-from discrete_optimization.rcpsp.rcpsp_model import RCPSPSolution
-from discrete_optimization.rcpsp.solver.cpm import CPM
+from discrete_optimization.rcpsp.problem import RcpspSolution
+from discrete_optimization.rcpsp.solvers.cpm import CpmRcpspSolver
 from scipy import stats
 from scipy.spatial import distance
 
@@ -172,7 +172,7 @@ def feature_precedence_done(
 
 
 def compute_cpm(do_domain):
-    cpm_solver = CPM(do_domain)
+    cpm_solver = CpmRcpspSolver(do_domain)
     path = cpm_solver.run_classic_cpm()
     cpm = cpm_solver.map_node
     cpm_esd = cpm[path[-1]]._ESD  # to normalize...
@@ -727,7 +727,7 @@ def evaluate_heuristic(self, individual, domains) -> float:
                 if normalized_values[i] not in {1, len(normalized_values)}
             ]
 
-            solution = RCPSPSolution(
+            solution = RcpspSolution(
                 problem=do_model,
                 rcpsp_permutation=normalized_values_for_do,
                 rcpsp_modes=modes,
@@ -941,7 +941,7 @@ def _get_next_action(
                 for j in scheduled_tasks_start_times
                 if j not in tasks_complete_dict
             }
-            solution = RCPSPSolution(
+            solution = RcpspSolution(
                 problem=do_model,
                 rcpsp_permutation=normalized_values_for_do,
                 rcpsp_modes=modes,
@@ -1111,7 +1111,7 @@ def _get_next_action(
 
         if run_sgs:
 
-            solution = RCPSPSolution(
+            solution = RcpspSolution(
                 problem=do_model,
                 rcpsp_permutation=normalized_values_for_do,
                 rcpsp_modes=modes,
@@ -1218,7 +1218,7 @@ def _get_next_action(
 
         if run_sgs:
 
-            solution = RCPSPSolution(
+            solution = RcpspSolution(
                 problem=do_model,
                 rcpsp_permutation=normalized_values_for_do,
                 rcpsp_modes=modes,