Files
Thesis/src/policies/PolicyEvaluator.py

407 lines
14 KiB
Python

from clearml import Task
from tqdm import tqdm
from src.policies.simple_baseline import BaselinePolicy
import pandas as pd
import numpy as np
import torch
import plotly.express as px
from functools import lru_cache
from src.utils.imbalance_price_calculator import ImbalancePriceCalculator
class PolicyEvaluator:
def __init__(self, baseline_policy: BaselinePolicy, task: Task = None):
self.baseline_policy = baseline_policy
self.ipc = ImbalancePriceCalculator(data_path="")
self.dates = baseline_policy.test_data["DateTime"].dt.date.unique()
self.dates = pd.to_datetime(self.dates)
### Load Imbalance Prices ###
imbalance_prices = pd.read_csv("data/imbalance_prices.csv", sep=";")
imbalance_prices["DateTime"] = pd.to_datetime(
imbalance_prices["DateTime"], utc=True
)
self.imbalance_prices = imbalance_prices.sort_values(by=["DateTime"])
self.penalties = [0, 1000, 1500]
self.profits = []
self.task = task
self.cache = {}
@lru_cache(maxsize=None)
def get_imbanlance_prices_for_date(self, date):
imbalance_prices_day = self.imbalance_prices[
self.imbalance_prices["DateTime"].dt.date == date
]
return imbalance_prices_day["Positive imbalance price"].values
def evaluate_for_date(
self,
date,
idx_samples,
test_loader,
charge_thresholds=np.arange(-1500, 1500, 50),
discharge_thresholds=np.arange(-1500, 1500, 50),
penalty: int = 0,
state_of_charge: float = 0.0,
):
if date in self.cache:
(reconstructed_imbalance_prices, real_imbalance_prices) = self.cache[date]
else:
idx = test_loader.dataset.get_idx_for_date(date.date())
if idx not in idx_samples:
print("No samples for idx: ", idx, date)
(initial, samples) = idx_samples[idx]
if len(initial.shape) == 2:
initial = initial.cpu().numpy()[0][-1]
else:
initial = initial.cpu().numpy()[-1]
samples = samples.cpu().numpy()
initial = np.repeat(initial, samples.shape[0])
combined = np.concatenate((initial.reshape(-1, 1), samples), axis=1)
reconstructed_imbalance_prices = (
self.ipc.get_imbalance_prices_2023_for_date_vectorized(date, combined)
)
reconstructed_imbalance_prices = torch.tensor(
reconstructed_imbalance_prices, device="cuda"
)
real_imbalance_prices = self.get_imbanlance_prices_for_date(date.date())
self.cache[date] = (reconstructed_imbalance_prices, real_imbalance_prices)
return self.profit_for_penalty(
reconstructed_imbalance_prices,
real_imbalance_prices,
penalty,
charge_thresholds,
discharge_thresholds,
state_of_charge=state_of_charge,
)
def optimize_penalty_for_target_charge_cycles(
self,
idx_samples,
test_loader,
initial_penalty,
target_charge_cycles,
initial_learning_rate=2,
max_iterations=10,
tolerance=10,
learning_rate_decay=0.9, # Factor to reduce the learning rate after each iteration
):
self.cache = {}
penalty = initial_penalty
learning_rate = initial_learning_rate
previous_gradient = None # Track the previous gradient to adjust learning rate based on progress
for iteration in range(max_iterations):
# Calculate profit and charge cycles for the current penalty
simulated_profit, simulated_charge_cycles = (
self.evaluate_test_set_for_penalty(idx_samples, test_loader, penalty)
)
print(
f"Iteration {iteration}: Penalty: {penalty}, Charge Cycles: {simulated_charge_cycles}, Profit: {simulated_profit}, Learning Rate: {learning_rate}"
)
# Calculate the gradient (difference) between the simulated and target charge cycles
gradient = simulated_charge_cycles - target_charge_cycles
# Optionally, adjust learning rate based on the change of gradient direction to avoid oscillation
if previous_gradient is not None and gradient * previous_gradient < 0:
learning_rate *= learning_rate_decay
# Update the penalty parameter in the direction of the gradient
penalty += (
learning_rate * gradient
) # Note: Using -= to move penalty in the opposite direction of gradient if necessary
# Update the previous gradient
previous_gradient = gradient
# Check if the charge cycles are close enough to the target
if abs(gradient) < tolerance:
print(f"Optimal penalty found after {iteration+1} iterations")
break
else:
print(
f"Reached max iterations ({max_iterations}) without converging to the target charge cycles"
)
# Re-calculate profit and charge cycles for the final penalty to return accurate results
profit, charge_cycles = self.evaluate_test_set_for_penalty(
idx_samples, test_loader, penalty
)
return penalty, profit, charge_cycles
def profit_for_penalty(
self,
reconstructed_imbalance_prices,
real_imbalance_prices,
penalty: int,
charge_thresholds,
discharge_thresholds,
state_of_charge=0.0,
):
"""_summary_
Args:
date (_type_): date to evaluate
reconstructed_imbalance_prices (_type_): predicted imbalance price
real_imbalance_prices (_type_): real imbalance price
penalty (int): penalty parameter to take into account
charge_thresholds (_type_): list of charge thresholds
discharge_thresholds (_type_): list of discharge thresholds
Returns:
_type_: returns the simulated profit, charge cycles, the found charge threshold and discharge threshold
"""
found_charge_thresholds, found_discharge_thresholds = (
self.baseline_policy.get_optimal_thresholds(
reconstructed_imbalance_prices,
charge_thresholds,
discharge_thresholds,
penalty,
battery_state_of_charge=state_of_charge,
)
)
predicted_charge_threshold = found_charge_thresholds.mean(axis=0)
predicted_discharge_threshold = found_discharge_thresholds.mean(axis=0)
### Determine Profits and Charge Cycles ###
simulated_profit, simulated_charge_cycles, new_state_of_charge = (
self.baseline_policy.simulate(
torch.tensor([[real_imbalance_prices]]),
torch.tensor([predicted_charge_threshold]),
torch.tensor([predicted_discharge_threshold]),
battery_state_of_charge=torch.tensor([state_of_charge]),
)
)
return (
simulated_profit[0][0].item(),
simulated_charge_cycles[0][0].item(),
predicted_charge_threshold.item(),
predicted_discharge_threshold.item(),
new_state_of_charge.squeeze(0).item(),
)
def evaluate_test_set(self, idx_samples, test_loader):
self.profits = []
self.cache = {}
for date in tqdm(self.dates):
try:
for penalty in self.penalties:
self.profits.append(
[
date,
penalty,
*self.evaluate_for_date(
date, idx_samples, test_loader, penalty=penalty
),
]
)
except KeyboardInterrupt:
print("Interrupted")
raise KeyboardInterrupt
except Exception as e:
print(e)
pass
self.profits = pd.DataFrame(
self.profits,
columns=[
"Date",
"Penalty",
"Profit",
"Charge Cycles",
"Charge Threshold",
"Discharge Threshold",
],
)
def evaluate_test_set_for_penalty(self, idx_samples, test_loader, penalty):
total_profit = 0
total_charge_cycles = 0
state_of_charge = 0.0
for date in tqdm(self.dates):
try:
profit, charge_cycles, _, _, new_state_of_charge = (
self.evaluate_for_date(
date,
idx_samples,
test_loader,
penalty=penalty,
state_of_charge=state_of_charge,
)
)
state_of_charge = new_state_of_charge
total_profit += profit
total_charge_cycles += charge_cycles
except KeyboardInterrupt:
print("Interrupted")
raise KeyboardInterrupt
except Exception as e:
print(e)
pass
return total_profit, total_charge_cycles
def plot_profits_table(self):
# Check if task or penalties are not set
if (
self.task is None
or not hasattr(self, "penalties")
or not hasattr(self, "profits")
):
print("Task, penalties, or profits not defined.")
return
if self.profits.empty:
print("Profits DataFrame is empty.")
return
# Aggregate profits and charge cycles by penalty, calculating totals and per-year values
aggregated = self.profits.groupby("Penalty").agg(
Total_Profit=("Profit", "sum"),
Total_Charge_Cycles=("Charge Cycles", "sum"),
Num_Days=("Date", "nunique"),
)
aggregated["Profit_Per_Year"] = (
aggregated["Total_Profit"] / aggregated["Num_Days"] * 365
)
aggregated["Charge_Cycles_Per_Year"] = (
aggregated["Total_Charge_Cycles"] / aggregated["Num_Days"] * 365
)
# Reset index to make 'Penalty' a column again and drop unnecessary columns
final_df = aggregated.reset_index().drop(
columns=["Total_Profit", "Total_Charge_Cycles", "Num_Days"]
)
# Rename columns to match expected output
final_df.columns = [
"Penalty",
"Total Profit (per year)",
"Total Charge Cycles (per year)",
]
# Profits till 400
profits_till_400 = self.get_profits_till_400()
# aggregate the final_df and profits_till_400 with columns: Penalty, total profit, total charge cycles, profit till 400, total charge cycles
final_df = final_df.merge(profits_till_400, on="Penalty")
# Log the final results table
self.task.get_logger().report_table(
"Test Set Results", "Profits per Penalty", iteration=0, table_plot=final_df
)
def plot_thresholds_per_day(self):
if self.task is None:
return
fig = px.line(
self.profits[self.profits["Penalty"] == 0],
x="Date",
y=["Charge Threshold", "Discharge Threshold"],
title="Charge and Discharge Thresholds per Day",
)
fig.update_layout(
width=1000,
height=600,
title_x=0.5,
)
self.task.get_logger().report_plotly(
"Thresholds per Day", "Thresholds per Day", iteration=0, figure=fig
)
def get_profits_as_scalars(self):
aggregated = self.profits.groupby("Penalty").agg(
Total_Profit=("Profit", "sum"),
Total_Charge_Cycles=("Charge Cycles", "sum"),
Num_Days=("Date", "nunique"),
)
aggregated["Profit_Per_Year"] = (
aggregated["Total_Profit"] / aggregated["Num_Days"] * 365
)
aggregated["Charge_Cycles_Per_Year"] = (
aggregated["Total_Charge_Cycles"] / aggregated["Num_Days"] * 365
)
# Reset index to make 'Penalty' a column again and drop unnecessary columns
final_df = aggregated.reset_index().drop(
columns=["Total_Profit", "Total_Charge_Cycles", "Num_Days"]
)
# Rename columns to match expected output
final_df.columns = ["Penalty", "Total Profit", "Total Charge Cycles"]
return final_df
def get_profits_till_400(self, profits: pd.DataFrame = None):
if profits is None:
profits = self.profits
# calculates profits until 400 charge cycles per year are reached
number_of_days = len(profits["Date"].unique())
usable_charge_cycles = (400 / 365) * number_of_days
# now sum the profit until the usable charge cycles are reached
penalty_profits = {}
penalty_charge_cycles = {}
for index, row in profits.iterrows():
penalty = row["Penalty"]
profit = row["Profit"]
charge_cycles = row["Charge Cycles"]
if penalty not in penalty_profits:
penalty_profits[penalty] = 0
penalty_charge_cycles[penalty] = 0
if penalty_charge_cycles[penalty] < usable_charge_cycles:
penalty_profits[penalty] += profit
penalty_charge_cycles[penalty] += charge_cycles
# transform profits to per year: penalty_profits / penalty_charge_cycles * 400 cycles per year
transformed_profits_per_year = {}
for penalty in penalty_profits:
transformed_profits_per_year[penalty] = (
penalty_profits[penalty] / penalty_charge_cycles[penalty] * 400
)
df = pd.DataFrame(
list(
zip(
penalty_profits.keys(),
penalty_profits.values(),
penalty_charge_cycles.values(),
transformed_profits_per_year.values(),
)
),
columns=[
"Penalty",
"Profit_till_400",
f"Cycles_till_400 (max {usable_charge_cycles})",
"Profit_per_year_till_400",
],
)
return df