# Copyright (c) 2019-2020, RTE (https://www.rte-france.com)
# See AUTHORS.txt
# This Source Code Form is subject to the terms of the Mozilla Public License, version 2.0.
# If a copy of the Mozilla Public License, version 2.0 was not distributed with this file,
# you can obtain one at http://mozilla.org/MPL/2.0/.
# SPDX-License-Identifier: MPL-2.0
# This file is part of Grid2Op, Grid2Op a testbed platform to model sequential decision making in power systems.
from typing import Optional, Union, Dict, Literal
import numpy as np
import hashlib
from datetime import datetime, timedelta
import grid2op
from grid2op.dtypes import dt_int
from grid2op.Chronics.gridValue import GridValue
from grid2op.Exceptions import ChronicsError
[docs]class FromNPY(GridValue):
"""
This class allows to generate some chronics compatible with grid2op if the data are provided in numpy format.
It also enables the use of the starting the chronics at different time than the original time and to end it before the end
of the chronics.
It is then much more flexible in its usage than the defaults chronics. But it is also much more error prone. For example, it does not check
the order of the loads / generators that you provide.
.. warning::
It assume the order of the elements are consistent with the powergrid backend ! It will not attempt to reorder the columns of the dataset
.. note::
The effect if "i_start" and "i_end" are persistant. If you set it once, it affects the object even after "env.reset()" is called.
If you want to modify them, you need to use the :func:`FromNPY.chronics.change_i_start` and :func:`FromNPY.chronics.change_i_end` methods
(and call `env.reset()`!)
TODO implement methods to change the loads / production "based on sampling" (online sampling instead of only reading data)
TODO implement the possibility to simulate maintenance / hazards "on the fly"
TODO implement hazards !
Examples
--------
Usage example, for what you don't really have to do:
.. code-block:: python
import grid2op
from grid2op.Chronics import FromNPY
# first retrieve the data that you want, the easiest wayt is to create an environment and read the data from it.
env_name = "l2rpn_case14_sandbox" # for example
env_ref = grid2op.make(env_name)
# retrieve the data
load_p = 1.0 * env_ref.chronics_handler.real_data.data.load_p
load_q = 1.0 * env_ref.chronics_handler.real_data.data.load_q
prod_p = 1.0 * env_ref.chronics_handler.real_data.data.prod_p
prod_v = 1.0 * env_ref.chronics_handler.real_data.data.prod_v
# now create an environment with these chronics:
env = grid2op.make(env_name,
chronics_class=FromNPY,
data_feeding_kwargs={"i_start": 5, # start at the "step" 5 NB first step is first observation, available with `obs = env.reset()`
"i_end": 18, # end index: data after that will not be considered (excluded as per python convention)
"load_p": load_p,
"load_q": load_q,
"prod_p": prod_p,
"prod_v": prod_v
## other parameters includes
# maintenance
# load_p_forecast
# load_q_forecast
# prod_p_forecast
# prod_v_forecast
# init_state # new in 1.10.2
})
# you can use env normally, including in runners
obs = env.reset()
# obs.load_p is load_p[5] (because you set "i_start" = 5, by default it's 0)
You can, after creation, change the data with:
.. code-block:: python
# create env as above
# retrieve some new values that you would like
new_load_p = ...
new_load_q = ...
new_prod_p = ...
new_prod_v = ...
# change the values
env.chronics_handler.real_data.change_chronics(new_load_p, new_load_q, new_prod_p, new_prod_v)
obs = env.reset() # mandatory if you want the change to be taken into account
# obs.load_p is new_load_p[5] (or rather load_p[env.chronics_handler.real_data._i_start])
.. seealso::
More usage examples in:
- :func:`FromNPY.change_chronics`
- :func:`FromNPY.change_forecasts`
- :func:`FromNPY.change_i_start`
- :func:`FromNPY.change_i_end`
Attributes
----------
TODO
"""
MULTI_CHRONICS = False
def __init__(
self,
load_p: np.ndarray,
load_q: np.ndarray,
prod_p: np.ndarray,
prod_v: Optional[np.ndarray] = None,
hazards: Optional[np.ndarray] = None,
maintenance: Optional[np.ndarray] = None,
load_p_forecast: Optional[np.ndarray] = None, # TODO forecasts !!
load_q_forecast: Optional[np.ndarray] = None,
prod_p_forecast: Optional[np.ndarray] = None,
prod_v_forecast: Optional[np.ndarray] = None,
time_interval: timedelta = timedelta(minutes=5),
max_iter: int = -1,
start_datetime: datetime = datetime(year=2019, month=1, day=1),
chunk_size: Optional[int] = None,
i_start: Optional[int] = None,
i_end: Optional[int] = None, # excluded, as always in python
init_state: Optional["grid2op.Action.BaseAction"] = None,
**kwargs
):
GridValue.__init__(
self,
time_interval=time_interval,
max_iter=max_iter,
start_datetime=start_datetime,
chunk_size=chunk_size,
)
self._i_start: int = i_start if i_start is not None else 0
self.__new_istart: Optional[int] = i_start
self.n_gen: int = prod_p.shape[1]
self.n_load: int = load_p.shape[1]
self.n_line: Union[int, None] = None
self._load_p: np.ndarray = 1.0 * load_p
self._load_q: np.ndarray = 1.0 * load_q
self._prod_p: np.ndarray = 1.0 * prod_p
self._prod_v = None
if prod_v is not None:
self._prod_v = 1.0 * prod_v
self.__new_load_p: Optional[np.ndarray] = None
self.__new_prod_p: Optional[np.ndarray] = None
self.__new_prod_v: Optional[np.ndarray] = None
self.__new_load_q: Optional[np.ndarray] = None
self._i_end: int = i_end if i_end is not None else load_p.shape[0]
self.__new_iend: Optional[int] = i_end
self.has_maintenance = False
self.maintenance = None
self.maintenance_duration = None
self.maintenance_time = None
if maintenance is not None:
self.has_maintenance = True
self.n_line = maintenance.shape[1]
assert load_p.shape[0] == maintenance.shape[0]
self.maintenance = maintenance # TODO copy
self.maintenance_time = (
np.zeros(shape=(self.maintenance.shape[0], self.n_line), dtype=dt_int)
- 1
)
self.maintenance_duration = np.zeros(
shape=(self.maintenance.shape[0], self.n_line), dtype=dt_int
)
for line_id in range(self.n_line):
self.maintenance_time[:, line_id] = self.get_maintenance_time_1d(
self.maintenance[:, line_id]
)
self.maintenance_duration[
:, line_id
] = self.get_maintenance_duration_1d(self.maintenance[:, line_id])
self.has_hazards = False
self.hazards = None
self.hazard_duration = None
if hazards is not None:
raise ChronicsError(
"This feature is not available at the moment. Fill a github issue at "
"https://github.com/Grid2Op/grid2op/issues/new?assignees=&labels=enhancement&template=feature_request.md&title="
)
self._forecasts = None
if load_p_forecast is not None:
assert load_q_forecast is not None
assert prod_p_forecast is not None
self._forecasts = FromNPY(
load_p=load_p_forecast,
load_q=load_q_forecast,
prod_p=prod_p_forecast,
prod_v=prod_v_forecast,
load_p_forecast=None,
load_q_forecast=None,
prod_p_forecast=None,
prod_v_forecast=None,
i_start=i_start,
i_end=i_end,
)
elif load_q_forecast is not None:
raise ChronicsError(
"if load_q_forecast is not None, then load_p_forecast should not be None"
)
elif prod_p_forecast is not None:
raise ChronicsError(
"if prod_p_forecast is not None, then load_p_forecast should not be None"
)
self._init_state = init_state
self._max_iter = min(self._i_end - self._i_start, load_p.shape[0])
[docs] def initialize(
self,
order_backend_loads,
order_backend_prods,
order_backend_lines,
order_backend_subs,
names_chronics_to_backend=None,
):
assert len(order_backend_prods) == self.n_gen
assert len(order_backend_loads) == self.n_load
if self.n_line is None:
self.n_line = len(order_backend_lines)
else:
assert len(order_backend_lines) == self.n_line
if self._forecasts is not None:
self._forecasts.initialize(
order_backend_loads,
order_backend_prods,
order_backend_lines,
order_backend_subs,
names_chronics_to_backend,
)
self.maintenance_time_nomaint = np.zeros(shape=(self.n_line,), dtype=dt_int) - 1
self.maintenance_duration_nomaint = np.zeros(shape=(self.n_line,), dtype=dt_int)
self.hazard_duration_nohaz = np.zeros(shape=(self.n_line,), dtype=dt_int)
self.curr_iter = 0
self.current_index = self._i_start - 1
self._max_iter = self._i_end - self._i_start
def _get_long_hash(self, hash_: hashlib.blake2b = None):
# get the "long hash" from blake2b
if hash_ is None:
hash_ = (
hashlib.blake2b()
) # should be faster than md5 ! (and safer, but we only care about speed here)
hash_.update(self._load_p.tobytes())
hash_.update(self._load_q.tobytes())
hash_.update(self._prod_p.tobytes())
if self._prod_v is not None:
hash_.update(self._prod_v.tobytes())
if self.maintenance is not None:
hash_.update(self.maintenance.tobytes())
if self.hazards is not None:
hash_.update(self.hazards.tobytes())
if self._forecasts:
self._forecasts._get_long_hash(hash_)
return hash_.digest()
[docs] def get_id(self) -> str:
"""
To return a unique ID of the chronics, we use a hash function (black2b), but it outputs a name too big (64 characters or so).
So we hash it again with md5 to get a reasonable length id (32 characters)
Returns:
str: the hash of the arrays (load_p, load_q, etc.) in the chronics
"""
long_hash_byte = self._get_long_hash()
# now shorten it with md5
short_hash = hashlib.md5(long_hash_byte)
return short_hash.hexdigest()
@staticmethod
def _create_dict_inj(res, obj_with_inj_data):
dict_ = {}
prod_v = None
if obj_with_inj_data._load_p is not None:
dict_["load_p"] = 1.0 * obj_with_inj_data._load_p[obj_with_inj_data.current_index, :]
if obj_with_inj_data._load_q is not None:
dict_["load_q"] = 1.0 * obj_with_inj_data._load_q[obj_with_inj_data.current_index, :]
array_gen_p = obj_with_inj_data._gen_p if hasattr(obj_with_inj_data, "_gen_p") else obj_with_inj_data._prod_p
if array_gen_p is not None:
dict_["prod_p"] = 1.0 * array_gen_p[obj_with_inj_data.current_index, :]
array_gen_v = obj_with_inj_data._gen_v if hasattr(obj_with_inj_data, "_gen_v") else obj_with_inj_data._prod_v
if array_gen_v is not None:
prod_v = 1.0 * array_gen_v[obj_with_inj_data.current_index, :]
if dict_:
res["injection"] = dict_
return prod_v
@staticmethod
def _create_dict_maintenance_hazards(res, obj_with_inj_data):
if obj_with_inj_data.maintenance is not None and obj_with_inj_data.has_maintenance:
res["maintenance"] = obj_with_inj_data.maintenance[obj_with_inj_data.current_index, :]
if obj_with_inj_data.hazards is not None and obj_with_inj_data.has_hazards:
res["hazards"] = obj_with_inj_data.hazards[obj_with_inj_data.current_index, :]
if (
obj_with_inj_data.maintenance_time is not None
and obj_with_inj_data.maintenance_duration is not None
and obj_with_inj_data.has_maintenance
):
maintenance_time = dt_int(1 * obj_with_inj_data.maintenance_time[obj_with_inj_data.current_index, :])
maintenance_duration = dt_int(
1 * obj_with_inj_data.maintenance_duration[obj_with_inj_data.current_index, :]
)
else:
maintenance_time = obj_with_inj_data.maintenance_time_nomaint
maintenance_duration = obj_with_inj_data.maintenance_duration_nomaint
if obj_with_inj_data.hazard_duration is not None and obj_with_inj_data.has_hazards:
hazard_duration = 1 * obj_with_inj_data.hazard_duration[obj_with_inj_data.current_index, :]
else:
hazard_duration = obj_with_inj_data.hazard_duration_nohaz
return maintenance_time, maintenance_duration, hazard_duration
[docs] def load_next(self):
self.current_index += 1
if (
self.current_index > self._i_end
or self.current_index >= self._load_p.shape[0]
):
raise StopIteration
res = {}
prod_v = FromNPY._create_dict_inj(res, self)
maintenance_time, maintenance_duration, hazard_duration = FromNPY._create_dict_maintenance_hazards(res, self)
self.current_datetime += self.time_interval
self.curr_iter += 1
return (
self.current_datetime,
res,
maintenance_time,
maintenance_duration,
hazard_duration,
prod_v,
)
[docs] def check_validity(
self, backend: Optional["grid2op.Backend.backend.Backend"]
) -> None:
# TODO raise the proper errors from ChronicsError here rather than AssertError
assert self._load_p.shape[0] == self._load_q.shape[0]
assert self._load_p.shape[0] == self._prod_p.shape[0]
if self._prod_v is not None:
assert self._load_p.shape[0] == self._prod_v.shape[0]
if self.hazards is not None:
assert self.hazards.shape[1] == self.n_line
if self.maintenance is not None:
assert self.maintenance.shape[1] == self.n_line
if self.maintenance_duration is not None:
assert self.n_line == self.maintenance_duration.shape[1]
if self.maintenance_time is not None:
assert self.n_line == self.maintenance_time.shape[1]
# TODO forecast
if self._forecasts is not None:
assert self._forecasts.n_line == self.n_line
assert self._forecasts.n_gen == self.n_gen
assert self._forecasts.n_load == self.n_load
assert self._load_p.shape[0] == self._forecasts._load_p.shape[0]
assert self._load_q.shape[0] == self._forecasts._load_q.shape[0]
assert self._prod_p.shape[0] == self._forecasts._prod_p.shape[0]
if self._prod_v is not None and self._forecasts._prod_v is not None:
assert self._prod_v.shape[0] == self._forecasts._prod_v.shape[0]
self._forecasts.check_validity(backend=backend)
[docs] def next_chronics(self):
# restart the chronics: read it again !
self.current_datetime = self.start_datetime
self.curr_iter = 0
if self.__new_istart is not None:
self._i_start = self.__new_istart
else:
self._i_start = 0
self.current_index = self._i_start
if self.__new_load_p is not None:
self._load_p = self.__new_load_p
self.__new_load_p = None
if self.__new_load_q is not None:
self._load_q = self.__new_load_q
self.__new_load_q = None
if self.__new_prod_p is not None:
self._prod_p = self.__new_prod_p
self.__new_prod_p = None
if self.__new_prod_v is not None:
self._prod_v = self.__new_prod_v
self.__new_prod_v = None
if self.__new_iend is None:
self._i_end = self._load_p.shape[0]
else:
self._i_end = self.__new_iend
if self._forecasts is not None:
# update the forecast
self._forecasts.next_chronics()
self.check_validity(backend=None)
self._max_iter = self._i_end - self._i_start
[docs] def done(self):
"""
INTERNAL
.. warning:: /!\\\\ Internal, do not use unless you know what you are doing /!\\\\
Compare to :func:`GridValue.done` an episode can be over for 2 main reasons:
- :attr:`GridValue.max_iter` has been reached
- There are no data in the numpy array.
- i_end has been reached
The episode is done if one of the above condition is met.
Returns
-------
res: ``bool``
Whether the episode has reached its end or not.
"""
res = False
if (
self.current_index >= self._i_end
or self.current_index >= self._load_p.shape[0]
):
res = True
elif self._max_iter > 0:
if self.curr_iter > self._max_iter:
res = True
return res
[docs] def forecasts(self):
"""
By default, forecasts are only made 1 step ahead.
We could change that. Do not hesitate to make a feature request
(https://github.com/Grid2Op/grid2op/issues/new?assignees=&labels=enhancement&template=feature_request.md&title=) if that is necessary for you.
"""
if self._forecasts is None:
return []
self._forecasts.current_index = self.current_index - 1
dt, dict_, *rest = self._forecasts.load_next()
return [(self.current_datetime + self.time_interval, dict_)]
[docs] def change_chronics(
self,
new_load_p: np.ndarray = None,
new_load_q: np.ndarray = None,
new_prod_p: np.ndarray = None,
new_prod_v: np.ndarray = None,
):
"""
Allows to change the data used by this class.
.. warning::
This has an effect only after "env.reset" has been called !
Args:
new_load_p (np.ndarray, optional): change the load_p. Defaults to None (= do not change).
new_load_q (np.ndarray, optional): change the load_q. Defaults to None (= do not change).
new_prod_p (np.ndarray, optional): change the prod_p. Defaults to None (= do not change).
new_prod_v (np.ndarray, optional): change the prod_v. Defaults to None (= do not change).
Examples
---------
.. code-block:: python
import grid2op
from grid2op.Chronics import FromNPY
# create an environment as in this class description (in short: )
load_p = ... # find somehow a suitable "load_p" array: rows represent time, columns the individual load
load_q = ...
prod_p = ...
prod_v = ...
# now create an environment with these chronics:
env = grid2op.make(env_name,
chronics_class=FromNPY,
data_feeding_kwargs={"load_p": load_p,
"load_q": load_q,
"prod_p": prod_p,
"prod_v": prod_v}
)
obs = env.reset() # obs.load_p is load_p[0] (or rather load_p[env.chronics_handler.real_data._i_start])
new_load_p = ... # find somehow a new suitable "load_p"
new_load_q = ...
new_prod_p = ...
new_prod_v = ...
env.chronics_handler.real_data.change_chronics(new_load_p, new_load_q, new_prod_p, new_prod_v)
# has no effect at this stage
obs = env.reset() # now has some effect !
# obs.load_p is new_load_p[0] (or rather load_p[env.chronics_handler.real_data._i_start])
"""
if new_load_p is not None:
self.__new_load_p = 1.0 * new_load_p
if new_load_q is not None:
self.__new_load_q = 1.0 * new_load_q
if new_prod_p is not None:
self.__new_prod_p = 1.0 * new_prod_p
if new_prod_v is not None:
self.__new_prod_v = 1.0 * new_prod_v
[docs] def change_forecasts(
self,
new_load_p: np.ndarray = None,
new_load_q: np.ndarray = None,
new_prod_p: np.ndarray = None,
new_prod_v: np.ndarray = None,
):
"""
Allows to change the data used by this class in the "obs.simulate" function.
.. warning::
This has an effect only after "env.reset" has been called !
Args:
new_load_p (np.ndarray, optional): change the load_p_forecast. Defaults to None (= do not change).
new_load_q (np.ndarray, optional): change the load_q_forecast. Defaults to None (= do not change).
new_prod_p (np.ndarray, optional): change the prod_p_forecast. Defaults to None (= do not change).
new_prod_v (np.ndarray, optional): change the prod_v_forecast. Defaults to None (= do not change).
Examples
---------
.. code-block:: python
import grid2op
from grid2op.Chronics import FromNPY
# create an environment as in this class description (in short: )
load_p = ... # find somehow a suitable "load_p" array: rows represent time, columns the individual load
load_q = ...
prod_p = ...
prod_v = ...
load_p_forecast = ...
load_q_forecast = ...
prod_p_forecast = ...
prod_v_forecast = ...
env = grid2op.make(env_name,
chronics_class=FromNPY,
data_feeding_kwargs={"load_p": load_p,
"load_q": load_q,
"prod_p": prod_p,
"prod_v": prod_v,
"load_p_forecast": load_p_forecast
"load_q_forecast": load_q_forecast
"prod_p_forecast": prod_p_forecast
"prod_v_forecast": prod_v_forecast
})
new_load_p_forecast = ... # find somehow a new suitable "load_p"
new_load_q_forecast = ...
new_prod_p_forecast = ...
new_prod_v_forecast = ...
env.chronics_handler.real_data.change_forecasts(new_load_p_forecast, new_load_q_forecast, new_prod_p_forecast, new_prod_v_forecast)
# has no effect at this stage
obs = env.reset() # now has some effect !
sim_o, *_ = obs.simulate() # sim_o.load_p has the values of new_load_p_forecast[0]
"""
if self._forecasts is None:
raise ChronicsError(
"You cannot change the forecast for this chronics are there are no forecasts enabled"
)
self._forecasts.change_chronics(
new_load_p=new_load_p,
new_load_q=new_load_q,
new_prod_p=new_prod_p,
new_prod_v=new_prod_v,
)
[docs] def max_timestep(self):
if self._max_iter >= 0:
return min(self._max_iter, self._load_p.shape[0], self._i_end)
return min(self._load_p.shape[0], self._i_end)
[docs] def change_i_start(self, new_i_start: Union[int, None]):
"""
Allows to change the "i_start".
.. warning::
It has only an affect after "env.reset()" is called.
Examples
--------
.. code-block:: python
import grid2op
from grid2op.Chronics import FromNPY
# create an environment as in this class description (in short: )
load_p = ... # find somehow a suitable "load_p" array: rows represent time, columns the individual load
load_q = ...
prod_p = ...
prod_v = ...
# now create an environment with these chronics:
env = grid2op.make(env_name,
chronics_class=FromNPY,
data_feeding_kwargs={"load_p": load_p,
"load_q": load_q,
"prod_p": prod_p,
"prod_v": prod_v}
)
obs = env.reset() # obs.load_p is load_p[0] (or rather load_p[env.chronics_handler.real_data._i_start])
env.chronics_handler.real_data.change_i_start(10)
obs = env.reset() # obs.load_p is load_p[10]
# indeed `env.chronics_handler.real_data._i_start` has been changed to 10.
# to undo all changes (and use the defaults) you can:
# env.chronics_handler.real_data.change_i_start(None)
"""
if new_i_start is not None:
self.__new_istart = int(new_i_start)
else:
self.__new_istart = None
[docs] def change_i_end(self, new_i_end: Union[int, None]):
"""
Allows to change the "i_end".
.. warning::
It has only an affect after "env.reset()" is called.
Examples
--------
.. code-block:: python
import grid2op
from grid2op.Chronics import FromNPY
# create an environment as in this class description (in short: )
load_p = ... # find somehow a suitable "load_p" array: rows represent time, columns the individual load
load_q = ...
prod_p = ...
prod_v = ...
# now create an environment with these chronics:
env = grid2op.make(env_name,
chronics_class=FromNPY,
data_feeding_kwargs={"load_p": load_p,
"load_q": load_q,
"prod_p": prod_p,
"prod_v": prod_v}
)
obs = env.reset()
env.chronics_handler.real_data.change_i_end(150)
obs = env.reset()
# indeed `env.chronics_handler.real_data._i_end` has been changed to 10.
# scenario lenght will be at best 150 !
# to undo all changes (and use the defaults) you can:
# env.chronics_handler.real_data.change_i_end(None)
"""
if new_i_end is not None:
self.__new_iend = int(new_i_end)
else:
self.__new_iend = None
[docs] def get_init_action(self, names_chronics_to_backend: Optional[Dict[Literal["loads", "prods", "lines"], Dict[str, str]]]=None) -> Union["grid2op.Action.playableAction.PlayableAction", None]:
# names_chronics_to_backend is ignored, names should be consistent between the environment
# and the initial state
return self._init_state