Merge pull request #609 from BDonnot/master

Upgrade to 1.10.2.dev3

Author: BDonnot

.gitignore

@@ -408,6 +408,8 @@ grid2op/tests/20240429_failed_tests_small.txt
 grid2op/tests/20240429_teq_test.txt
 grid2op/tests/req_38_np121
 test_make_2_envs.py
+getting_started/env_py38_grid2op110_ray110.ipynb
+getting_started/env_py38_grid2op110_ray210.ipynb
 
 # profiling files
 **.prof

CHANGELOG.rst

@@ -52,6 +52,9 @@ Change Log
   and `FromChronix2grid` are not supported at the moment.
 - [ADDED] an "Handler" (`JSONInitStateHandler`) that can set the grid to an initial state (so as to make
   compatible the `FromHandlers` time series class with this new feature)
+- [ADDED] some more type hints in the `GridObject` class 
+- [ADDED] Possibility to deactive the support of shunts if subclassing `PandaPowerBackend`
+  (and add some basic tests)
 - [FIXED] a small issue that could lead to having 
   "redispatching_unit_commitment_availble" flag set even if the redispatching
   data was not loaded correctly
@@ -68,6 +71,8 @@ Change Log
   and most of the time incorrectly)
 - [FIXED] on `RemoteEnv` class (impact all multi process environment): the kwargs used to build then backend
   where not used which could lead to"wrong" backends being used in the sub processes.
+- [FIXED] a bug when the name of the times series and the names of the elements in the backend were 
+  different: it was not possible to set `names_chronics_to_grid` correctly when calling `env.make`
 - [IMPROVED] documentation about `obs.simulate` to make it clearer the 
   difference between env.step and obs.simulate on some cases
 - [IMPROVED] type hints on some methods of `GridObjects`
@@ -79,6 +84,8 @@ Change Log
 - [IMPROVED] the way the "grid2op compat" mode is handled
 - [IMPROVED] the coverage of the tests in the "test_basic_env_ls.py" to test more in depth lightsim2grid
   (creation of multiple environments, grid2op compatibility mode)
+- [IMPROVED] the function to test the backend interface in case when shunts are not supported 
+  (improved test `AAATestBackendAPI.test_01load_grid`)
 
 [1.10.1] - 2024-03-xx
 ----------------------

docs/conf.py

@@ -22,7 +22,7 @@
 author = 'Benjamin Donnot'
 
 # The full version, including alpha/beta/rc tags
-release = '1.10.2.dev2'
+release = '1.10.2.dev3'
 version = '1.10'
 
 

docs/environment.rst

@@ -104,6 +104,58 @@ increase the training time, especially at the beginning. This is due to the fact
 `env.reset` is called, the whole chronics is read from the hard drive. If you want to lower this
 impact then you might consult the :ref:`environment-module-data-pipeline` page of the doc.
 
+Go to the next scenario
+++++++++++++++++++++++++
+
+Starting grid2op 1.9.8 we attempt to make an easier user experience in the
+selection of time series, seed, initial state of the grid, etc.
+
+All of the above can be done when calling `env.reset()` function.
+
+For customizing the seed, you can for example do:
+
+.. code-block:: python
+
+    import grid2op
+    env_name = "l2rpn_case14_sandbox"
+    env = grid2op.make(env_name)
+
+    obs = env.reset(seed=0)
+
+For customizing the time series id you want to use:
+
+.. code-block:: python
+
+    import grid2op
+    env_name = "l2rpn_case14_sandbox"
+    env = grid2op.make(env_name)
+
+    obs = env.reset(options={"time serie id": 1})  # time serie by id (sorted alphabetically)
+    # or
+    obs = env.reset(options={"time serie id": "0001"})  # time serie by name (folder name)
+
+For customizing the initial state of the grid, for example forcing the
+powerline 0 to be disconnected in the initial observation:
+
+.. code-block:: python
+
+    import grid2op
+    env_name = "l2rpn_case14_sandbox"
+    env = grid2op.make(env_name)
+
+    init_state_dict = {"set_line_status": [(0, -1)]}
+    obs = env.reset(options={"init state": init_state_dict})
+
+
+Feel free to consult the documentation of the :func:`Environment.reset` function
+for more information (this doc might be outdated, the one of the function should 
+be more up to date with the code).
+
+.. info::
+    In the near future (next few releases) we will also attempt to make the 
+    customization of the `parameters` or the `skip number of steps`, `maximum duration 
+    of the scenarios` also available in `env.reset()` options.
+
 .. _environment-module-chronics-info:
 
 Time series Customization
@@ -141,10 +193,15 @@ the call to "env.reset". This gives the following code:
     # and now the loop starts
     for i in range(episode_count):
         ###################################
-        env.set_id(THE_CHRONIC_ID)
+        # with recent grid2op 
+        obs = env.reset(options={"time serie id": THE_CHRONIC_ID})
         ###################################
 
-        obs = env.reset()
+        ###################################
+        # 'old method (oldest grid2op version)'
+        # env.set_id(THE_CHRONIC_ID)
+        # obs = env.reset()
+        ###################################
 
         # now play the episode as usual
         while True:

getting_started/11_IntegrationWithExistingRLFrameworks.ipynb

@@ -135,6 +135,37 @@
     "\n",
     "More information are provided here: https://grid2op.readthedocs.io/en/latest/environment.html#splitting-into-raining-validation-test-scenarios\n",
     "\n",
+    "### Use the `experimental_read_from_local_dir` flag\n",
+    "\n",
+    "This flag allows python to better \"understands\" the classes in grid2op and avoid lots of issue with pickle / multi processing etc.\n",
+    "\n",
+    "The complete documentation is available here https://grid2op.readthedocs.io/en/latest/environment.html#grid2op.Environment.BaseEnv.generate_classes\n",
+    "\n",
+    "Basically, once, and only once, outside of this process, you can call:\n",
+    "\n",
+    "```python\n",
+    "import grid2op\n",
+    "env_name = \"l2rpn_case14_sandbox\"  # or any other name\n",
+    "\n",
+    "env = grid2op.make(env_name, ...)  # again: redo this step each time you customize \"...\"\n",
+    "# for example if you change the `action_class` or the `backend` etc.\n",
+    "\n",
+    "env.generate_classes()\n",
+    "```\n",
+    "\n",
+    "Then, each time you want to reload the same environment, you can do:\n",
+    "\n",
+    "```python\n",
+    "import grid2op\n",
+    "env_name = SAME NAME AS ABOVE\n",
+    "env = grid2op.make(env_name,\n",
+    "                   experimental_read_from_local_dir=True,\n",
+    "                   ..., # SAME ENV CUSTOMIZATION AS ABOVE\n",
+    "                  )\n",
+    "```\n",
+    "\n",
+    "This is known to solve bug related to multi processing, but also to reduce the amount of RAM taken (in some cases) as well as creation time (in some cases)\n",
+    "\n",
     "### Other steps\n",
     "\n",
     "The grid2op documentation is full of details to \"optimize\" the number of steps you can do per seconds. This number can rise from a few dozen per seconds to around a thousands per seconds with proper care.\n",
@@ -190,25 +221,13 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "import gym\n",
+    "import gymnasium\n",
     "import numpy as np\n",
     "from grid2op.gym_compat import GymEnv\n",
     "env_gym_init = GymEnv(env_glop)\n",
     "env_gym = GymEnv(env_glop)\n",
-    "print(f\"The \\\"env_gym\\\" is a gym environment: {isinstance(env_gym, gym.Env)}\")\n",
-    "obs_gym = env_gym.reset()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "<font size=\"3\" color=\"red\"> In this notebook, we only present some basic (and really \"detailed\" use of the `GymEnv`). </font>\n",
-    "    \n",
-    "<font size=\"3\" color=\"red\"> This is especially suited for advanced users wanting a deep control over everything happening. </font>\n",
-    "\n",
-    "<font size=\"3\" color=\"red\"> For a less advanced usage, feel free to consult the l2rpn baselines package, that embed some usefull environments, compatible with gym, that can embed some heuristics and other \"quality of life\" features. Feel free to use the l2rpn baselines package for more information.\n",
-    "</font>"
+    "print(f\"The \\\"env_gym\\\" is a gym environment: {isinstance(env_gym, gymnasium.Env)}\")\n",
+    "obs_gym, info = env_gym.reset()"
    ]
   },
   {
@@ -388,7 +407,7 @@
    "outputs": [],
    "source": [
     "from grid2op.gym_compat import ScalerAttrConverter\n",
-    "from gym.spaces import Box\n",
+    "from gymnasium.spaces import Box\n",
     "ob_space = env_gym.observation_space\n",
     "ob_space = ob_space.reencode_space(\"actual_dispatch\",\n",
     "                                   ScalerAttrConverter(substract=0.,\n",
@@ -518,11 +537,11 @@
    "source": [
     "# gym specific, we simply do a copy paste of what we did in the previous cells, wrapping it in the\n",
     "# MyEnv class, and train a Proximal Policy Optimisation based agent\n",
-    "import gym\n",
+    "import gymnasium\n",
     "import ray\n",
     "import numpy as np\n",
     "      \n",
-    "class MyEnv(gym.Env):\n",
+    "class MyEnv(gymnasium.Env):\n",
     "    def __init__(self, env_config):\n",
     "        import grid2op\n",
     "        from grid2op.gym_compat import GymEnv\n",
@@ -537,7 +556,7 @@
     "\n",
     "        # 2. create the gym environment\n",
     "        self.env_gym = GymEnv(self.env_glop)\n",
-    "        obs_gym = self.env_gym.reset()\n",
+    "        obs_gym, info = self.env_gym.reset()\n",
     "\n",
     "        # 3. (optional) customize it (see section above for more information)\n",
     "        ## customize action space\n",
@@ -576,9 +595,9 @@
     "        # 4. bis: to avoid other type of issues, we recommend to build the action space and observation\n",
     "        # space directly from the spaces class.\n",
     "        d = {k: v for k, v in self.env_gym.observation_space.spaces.items()}\n",
-    "        self.observation_space = gym.spaces.Dict(d)\n",
+    "        self.observation_space = gymnasium.spaces.Dict(d)\n",
     "        a = {k: v for k, v in self.env_gym.action_space.items()}\n",
-    "        self.action_space = gym.spaces.Dict(a)\n",
+    "        self.action_space = gymnasium.spaces.Dict(a)\n",
     "\n",
     "    def reset(self):\n",
     "        obs = self.env_gym.reset()\n",
@@ -791,7 +810,7 @@
     "                                                      )\n",
     "                                                 }\n",
     "                                         )\n",
-    "obs_gym = env_sb.reset()"
+    "obs_gym, info = env_sb.reset()"
    ]
   },
   {
@@ -877,13 +896,13 @@
    "outputs": [],
    "source": [
     "from grid2op.gym_compat import BoxGymActSpace\n",
-    "scale_gen =  env_sb.init_env.gen_max_ramp_up + env_sb.init_env.gen_max_ramp_down\n",
-    "scale_gen[~env_sb.init_env.gen_redispatchable] = 1.0\n",
+    "scaler_gen =  env_sb.init_env.gen_max_ramp_up + env_sb.init_env.gen_max_ramp_down\n",
+    "scaler_gen = scaler_gen[env_sb.init_env.gen_redispatchable]\n",
     "env_sb.action_space = BoxGymActSpace(env_sb.init_env.action_space,\n",
     "                                     attr_to_keep=[\"redispatch\"],\n",
-    "                                     multiply={\"redispatch\": scale_gen},\n",
+    "                                     multiply={\"redispatch\": scaler_gen},\n",
     "                                    )\n",
-    "obs_gym = env_sb.reset()"
+    "obs_gym, info = env_sb.reset()"
    ]
   },
   {
@@ -937,7 +956,7 @@
     "reencoded_act_space = MultiDiscreteActSpace(env_sb.init_env.action_space,\n",
     "                                           attr_to_keep=[\"set_line_status\", \"set_bus\", \"redispatch\"])\n",
     "env_sb.action_space = reencoded_act_space\n",
-    "obs_gym = env_sb.reset()"
+    "obs_gym, info = env_sb.reset()"
    ]
   },
   {
@@ -1041,7 +1060,7 @@
     "                                                      )\n",
     "                                                 }\n",
     "                                         )\n",
-    "obs_gym = env_tfa.reset()"
+    "obs_gym, info = env_tfa.reset()"
    ]
   },
   {
@@ -1297,7 +1316,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.10"
+   "version": "3.10.13"
   }
  },
  "nbformat": 4,