Skip to content

Commit

Permalink
Merge pull request #637 from BDonnot/dev-switches
Browse files Browse the repository at this point in the history
add more tests, fix issues when setting topology of multiple substations
  • Loading branch information
BDonnot authored Oct 11, 2024
2 parents 352749f + 45cac39 commit 24f3c3c
Showing 700 changed files with 26,242 additions and 366 deletions.
61 changes: 60 additions & 1 deletion .github/workflows/main.yml
Original file line number Diff line number Diff line change
@@ -203,7 +203,7 @@ jobs:
steps:

- name: Checkout sources
uses: actions/checkout@v1
uses: actions/checkout@v3
with:
submodules: true

@@ -231,6 +231,65 @@ jobs:
- name: Test the automatic generation of classes in the env folder
run: |
python -m unittest grid2op/tests/automatic_classes.py -f
detailed_topology:
name: Test detailed topology on ${{ matrix.config.name }}
runs-on: ${{ matrix.config.os }}
strategy:
matrix:
config:
# - {
# name: darwin,
# os: macos-latest,
# }
# - {
# name: windows,
# os: windows-2019,
# }
- {
name: ubuntu,
os: ubuntu-latest,
}
python:
- {
name: cp39,
version: '3.9',
}
- {
name: cp312,
version: '3.12',
}

steps:

- name: Checkout sources
uses: actions/checkout@v3
with:
submodules: true

- name: Setup Python
uses: actions/setup-python@v2
with:
python-version: ${{ matrix.python.version }}

- name: Install Python dependencies
run: |
python -m pip install --upgrade pip
python -m pip install --upgrade wheel
python -m pip install --upgrade setuptools
- name: Build wheel
run: python setup.py bdist_wheel

- name: Install wheel
shell: bash
run: |
python -m pip install dist/*.whl --user
pip freeze
- name: Test the working of the detailed topo representation
run: |
python -m unittest grid2op/tests/indep_test_detailed_topo_extended.py -f
package:
name: Test install
9 changes: 9 additions & 0 deletions .gitignore
Original file line number Diff line number Diff line change
@@ -416,6 +416,15 @@ getting_started/venv_310_ray/
grid2op/tests/venv_test_autoclass/
test_eduardo.py
grid2op/tests/failed_test*
grid2op/tests/ampl_from_iidm.py
grid2op/tests/do_I_repeat.txt
grid2op/tests/parse_res35_working.txt
grid2op/tests/parse_res_35 (copie).txt
grid2op/tests/parse_res_35.txt
grid2op/tests/res_algo_topo_35.txt
grid2op/tests/test_topo_ampl/
grid2op/tests/test_topo_ampl2/
venv_312

# profiling files
**.prof
8 changes: 8 additions & 0 deletions CHANGELOG.rst
Original file line number Diff line number Diff line change
@@ -75,6 +75,14 @@ Next release
- [FIXED] the correct `AmbiguousAction` is now raised when grid2op does not understand
what an action should be doing (an incorrect `IllegalAction` used to be sent)
- [FIXED] a test in `test_ActionProperties` did not test the correct property
- [FIXED] an error in the descirption of the `educ_case14_storage` environment
(wrong sign for the slack generator)
- [FIXED] the environment would not load in case of an incorrect "layout.json"
instead of raising a warning.
- [FIXED] some issue with gym_compat module for "newest" version of
gymnasium (1.0.0)
- [IMPROVED] error message when forecasts are not correctly set-up
- [IMPROVED] an error message when loading a grid with forecasts

[1.10.3] - 2024-07-12
-------------------------
2 changes: 1 addition & 1 deletion docs/conf.py
Original file line number Diff line number Diff line change
@@ -22,7 +22,7 @@
author = 'Benjamin Donnot'

# The full version, including alpha/beta/rc tags
release = '1.10.4.dev0'
release = '1.10.4.dev1'
version = '1.10'


62 changes: 31 additions & 31 deletions docs/gym.rst
Original file line number Diff line number Diff line change
@@ -29,20 +29,20 @@ your base code.

More information on the section :ref:`gymnasium_gym`

Before grid2op 1.2.0 only some classes fully implemented the open AI gym interface:
Before grid2op 1.2.0 only some classes fully implemented the gymnasium interface:

- the :class:`grid2op.Environment` (with methods such as `env.reset`, `env.step` etc.)
- the :class:`grid2op.Agent` (with the `agent.act` etc.)
- the creation of pre defined environments (with `grid2op.make`)


Starting from 1.2.0 we implemented some automatic converters that are able to automatically map
grid2op representation for the action space and the observation space into open AI gym "spaces". More precisely these
grid2op representation for the action space and the observation space into gymnasium "spaces". More precisely these
are represented as gym.spaces.Dict.

As of grid2op 1.4.0 we tighten the gap between openAI gym and grid2op by introducing the dedicated module
As of grid2op 1.4.0 we tighten the gap between gymnasium and grid2op by introducing the dedicated module
`grid2op.gym_compat` . Withing this module there are lots of functionalities to convert a grid2op environment
into a gym environment (that inherit `gym.Env` instead of "simply" implementing the open ai gym interface).
into a gymnasium environment (that inherit `gymnasium.Env` instead of "simply" implementing the gymnasium interface).


A simple usage is:
@@ -55,12 +55,12 @@ A simple usage is:
env_name = "l2rpn_case14_sandbox" # or any other grid2op environment name
g2op_env = grid2op.make(env_name) # create the gri2op environment
gym_env = GymEnv(g2op_env) # create the gym environment
gym_env = GymEnv(g2op_env) # create the gymnasium environment
# check that this is a properly defined gym environment:
# check that this is a properly defined gymnasium environment:
import gym
print(f"Is gym_env and open AI gym environment: {isinstance(gym_env, gym.Env)}")
# it shows "Is gym_env and open AI gym environment: True"
print(f"Is gym_env a gymnasium environment: {isinstance(gym_env, gym.Env)}")
# it shows "Is gym_env a gymnasium environment: True"
.. note::

@@ -73,9 +73,9 @@ A simple usage is:
.. warning::
The `gym` package has some breaking API change since its version 0.26. We attempted,
in grid2op, to maintain compatibility both with former versions and later ones. This makes **this
class behave differently depending on the version of gym you have installed** !
class behave differently depending on the version of gymnasium you have installed** !

The main changes involve the functions `env.step` and `env.reset` (core gym functions)
The main changes involve the functions `env.step` and `env.reset` (core gymnasium functions)

This page is organized as follow:

@@ -164,7 +164,7 @@ You can transform the observation space as you wish. There are some examples in

Default Action space
******************************
The default action space is also a type of gym Dict. As for the observation space above, it is a
The default action space is also a type of gymnasium Dict. As for the observation space above, it is a
straight translation from the attribute of the action to the key of the dictionary. This gives:

- "change_bus": MultiBinary(`env.dim_topo`)
@@ -177,7 +177,7 @@ straight translation from the attribute of the action to the key of the dictiona
- "raise_alarm": MultiBinary(`env.dim_alarms`)
- "raise_alert": MultiBinary(`env.dim_alerts`)

For example you can create a "gym action" (for the default encoding) like:
For example you can create a "gymnasium action" (for the default encoding) like:

.. code-block:: python
@@ -191,27 +191,27 @@ For example you can create a "gym action" (for the default encoding) like:
gym_env = GymEnv(env)
seed = ...
obs, info = gym_env.reset(seed) # for new gym interface
obs, info = gym_env.reset(seed) # for new gymnasium interface
# do nothing
gym_act = {}
obs, reward, done, truncated, info = gym_env.step(gym_act)
#change the bus of the element 6 and 7 of the "topo_vect"
gym_act = {}
gym_act["change_bus"] = np.zeros(env.dim_topo, dtype=np.int8) # gym encoding of a multi binary
gym_act["change_bus"] = np.zeros(env.dim_topo, dtype=np.int8) # gymnasium encoding of a multi binary
gym_act["change_bus"][[6, 7]] = 1
obs, reward, done, truncated, info = gym_env.step(gym_act)
# redispatch generator 2 of 1.7MW
gym_act = {}
gym_act["redispatch"] = np.zeros(env.n_gen, dtype=np.float32) # gym encoding of a Box
gym_act["redispatch"] = np.zeros(env.n_gen, dtype=np.float32) # gymnasium encoding of a Box
gym_act["redispatch"][2] = 1.7
obs, reward, done, truncated, info = gym_env.step(gym_act)
# set the bus of element 8 and 9 to bus 2
gym_act = {}
gym_act["set_bus"] = np.zeros(env.dim_topo, dtype=int) # gym encoding of a Box
gym_act["set_bus"] = np.zeros(env.dim_topo, dtype=int) # gymnasium encoding of a Box
gym_act["set_bus"][[8, 9]] = 2
obs, reward, done, truncated, info = gym_env.step(gym_act)
@@ -238,7 +238,7 @@ If you want a full control on this spaces, you need to implement something like:
env = grid2op.make(env_name)
from grid2op.gym_compat import GymEnv
# this of course will not work... Replace "AGymSpace" with a normal gym space, like Dict, Box, MultiDiscrete etc.
# this of course will not work... Replace "AGymSpace" with a normal gymnasium space, like Dict, Box, MultiDiscrete etc.
from gym.spaces import AGymSpace
gym_env = GymEnv(env)
@@ -253,7 +253,7 @@ If you want a full control on this spaces, you need to implement something like:
def to_gym(self, observation):
# this is this very same function that you need to implement
# it should have this exact name, take only one observation (grid2op) as input
# and return a gym object that belong to your space "AGymSpace"
# and return a gymnasium object that belong to your space "AGymSpace"
return SomethingThatBelongTo_AGymSpace
# eg. return np.concatenate((obs.gen_p * 0.1, np.sqrt(obs.load_p))
@@ -268,7 +268,7 @@ And for the action space:
env = grid2op.make(env_name)
from grid2op.gym_compat import GymEnv
# this of course will not work... Replace "AGymSpace" with a normal gym space, like Dict, Box, MultiDiscrete etc.
# this of course will not work... Replace "AGymSpace" with a normal gymnasium space, like Dict, Box, MultiDiscrete etc.
from gym.spaces import AGymSpace
gym_env = GymEnv(env)
@@ -282,7 +282,7 @@ And for the action space:
def from_gym(self, gym_action):
# this is this very same function that you need to implement
# it should have this exact name, take only one action (member of your gym space) as input
# it should have this exact name, take only one action (member of your gymnasium space) as input
# and return a grid2op action
return TheGymAction_ConvertedTo_Grid2op_Action
# eg. return np.concatenate((obs.gen_p * 0.1, np.sqrt(obs.load_p))
@@ -311,7 +311,7 @@ and divide input data by `divide`):
env_name = "l2rpn_case14_sandbox" # or any other grid2op environment name
g2op_env = grid2op.make(env_name) # create the gri2op environment
gym_env = GymEnv(g2op_env) # create the gym environment
gym_env = GymEnv(g2op_env) # create the gymnasium environment
ob_space = gym_env.observation_space
ob_space = ob_space.reencode_space("actual_dispatch",
@@ -336,7 +336,7 @@ the log of the loads instead of giving the direct value to your agent. This can
env_name = "l2rpn_case14_sandbox" # or any other grid2op environment name
g2op_env = grid2op.make(env_name) # create the gri2op environment
gym_env = GymEnv(g2op_env) # create the gym environment
gym_env = GymEnv(g2op_env) # create the gymnasium environment
ob_space = gym_env.observation_space
shape_ = (g2op_env.n_load, )
@@ -350,7 +350,7 @@ the log of the loads instead of giving the direct value to your agent. This can
)
gym_env.observation_space = ob_space
# and now you will get the key "log_load" as part of your gym observation.
# and now you will get the key "log_load" as part of your gymnasium observation.
A detailed list of such "converter" is documented on the section "Detailed Documentation by class". In
the table below we describe some of them (**nb** if you notice a converter is not displayed there,
@@ -360,11 +360,11 @@ do not hesitate to write us a "feature request" for the documentation, thanks in
Converter name Objective
============================================= ============================================================
:class:`ContinuousToDiscreteConverter` Convert a continuous space into a discrete one
:class:`MultiToTupleConverter` Convert a gym MultiBinary to a gym Tuple of gym Binary and a gym MultiDiscrete to a Tuple of Discrete
:class:`MultiToTupleConverter` Convert a gymnasium MultiBinary to a gymnasium Tuple of gymnasium Binary and a gymnasium MultiDiscrete to a Tuple of Discrete
:class:`ScalerAttrConverter` Allows to scale (divide an attribute by something and subtract something from it)
`BaseGymSpaceConverter.add_key`_ Allows you to compute another "part" of the observation space (you add an information to the gym space)
`BaseGymSpaceConverter.add_key`_ Allows you to compute another "part" of the observation space (you add an information to the gymnasium space)
`BaseGymSpaceConverter.keep_only_attr`_ Allows you to specify which part of the action / observation you want to keep
`BaseGymSpaceConverter.ignore_attr`_ Allows you to ignore some attributes of the action / observation (they will not be part of the gym space)
`BaseGymSpaceConverter.ignore_attr`_ Allows you to ignore some attributes of the action / observation (they will not be part of the gymnasium space)
============================================= ============================================================

.. warning::
@@ -383,7 +383,7 @@ Converter name Objective
.. note::

With the "converters" above, note that the observation space AND action space will still
inherit from gym Dict.
inherit from gymnasium Dict.

They are complex spaces that are not well handled by some RL framework.

@@ -395,19 +395,19 @@ Converter name Objective
Customizing the action and observation space, into Box or Discrete
*******************************************************************

The use of the converter above is nice if you can work with gym Dict, but in some cases, or for some frameworks
The use of the converter above is nice if you can work with gymnasium Dict, but in some cases, or for some frameworks
it is not convenient to do it at all.

TO alleviate this problem, we developed 4 types of gym action space, following the architecture
TO alleviate this problem, we developed 4 types of gymnasium action space, following the architecture
detailed in subsection :ref:`base_gym_space_function`

=============================== ============================================================
Converter name Objective
=============================== ============================================================
:class:`BoxGymObsSpace` Convert the observation space to a single "Box"
:class:`BoxGymActSpace` Convert a gym MultiBinary to a gym Tuple of gym Binary and a gym MultiDiscrete to a Tuple of Discrete
:class:`BoxGymActSpace` Convert a gymnasium MultiBinary to a gymnasium Tuple of gymnasium Binary and a gymnasium MultiDiscrete to a Tuple of Discrete
:class:`MultiDiscreteActSpace` Allows to scale (divide an attribute by something and subtract something from it)
:class:`DiscreteActSpace` Allows you to compute another "part" of the observation space (you add an information to the gym space)
:class:`DiscreteActSpace` Allows you to compute another "part" of the observation space (you add an information to the gymnasium space)
=============================== ============================================================

They can all be used like:
6 changes: 3 additions & 3 deletions docs/makeenv.rst
Original file line number Diff line number Diff line change
@@ -25,11 +25,11 @@ To get started with such an environment, you can simply do:
You can consult the different notebooks in the `getting_stared` directory of this package for more information on
how to use it.

Created Environment should behave exactly like a gym environment. If you notice any unwanted behavior, please address
Created Environment should behave exactly like a gymnasium environment. If you notice any unwanted behavior, please address
an issue in the official grid2op repository: `Grid2Op <https://github.com/rte-france/Grid2Op>`_

The environment created with this method should be fully compatible with the gym framework: if you are developing
a new algorithm of "Reinforcement Learning" and you used the openai gym framework to do so, you can port your code
The environment created with this method should be fully compatible with the gymnasium framework: if you are developing
a new algorithm of "Reinforcement Learning" and you used the openai gymnasium framework to do so, you can port your code
in a few minutes (basically this consists in adapting the input and output dimension of your BaseAgent) and make it work
with a Grid2Op environment. An example of such modifications is exposed in the getting_started/ notebooks.

2 changes: 1 addition & 1 deletion docs/model_free.rst
Original file line number Diff line number Diff line change
@@ -8,7 +8,7 @@ Model Free Reinforcement Learning

See some example in "l2rpn-baselines" package for now !

The main idea is first to convert the grid2op environment to a gym environment, for example using :ref:`openai-gym`.
The main idea is first to convert the grid2op environment to a gymnasium environment, for example using :ref:`openai-gym`.
And then use some libaries available,
for example `Stable Baselines <https://stable-baselines3.readthedocs.io/en/master/>`_ or
`RLLIB <https://docs.ray.io/en/latest/rllib/index.html>`_
8 changes: 4 additions & 4 deletions docs/plot.rst
Original file line number Diff line number Diff line change
@@ -76,10 +76,10 @@ An possible output will look like this:
Render the state of the grid
-----------------------------

During the gym loop
++++++++++++++++++++
During the gymnasium loop
++++++++++++++++++++++++++
In Grid2Op we also made available the possibility to render the state of the grid that your agent sees before taking
an action. This can be done with the provided environments following openAI gym interface like this:
an action. This can be done with the provided environments following gymnasium interface like this:

.. code-block:: python
@@ -104,7 +104,7 @@ significantly.

Offline, after the scenarios were played
++++++++++++++++++++++++++++++++++++++++
In Grid2Op, you can execute a :ref:`runner-module` to perform the "gym loops" and store the results
In Grid2Op, you can execute a :ref:`runner-module` to perform the "gymnasium loops" and store the results
in a standardized manner. Once stored, the results can be loaded back and "replayed" using the appropriate
class. Here is how you can do this:

Loading

0 comments on commit 24f3c3c

Please sign in to comment.