@@ -24,7 +24,7 @@
> [!IMPORTANT]
> **FinRL-X** is the next-generation evolution of FinRL, designed for AI-native, modular, and production-oriented quantitative trading.
->
+>
> - **This repository (`FinRL`)** preserves the original end-to-end educational and research framework.
> - **For the latest architecture, live trading deployment, and production-focused development, please use [`FinRL-X / FinRL-Trading`](https://github.com/AI4Finance-Foundation/FinRL-Trading).**
@@ -87,7 +87,7 @@ Key contributors include:
- [**Hongyang (Bruce) Yang**](https://www.linkedin.com/in/brucehy/) – research and development on financial reinforcement learning frameworks, market environments, and quantitative trading applications
- [other contributors…]
-
+
## Overview
FinRL is the original open-source framework for financial reinforcement learning, organized around three core layers:
diff --git a/examples/FinRL_StockTrading_2026_1_data.py b/examples/FinRL_StockTrading_2026_1_data.py
index b40f165ac4..7c0ca5e01c 100644
--- a/examples/FinRL_StockTrading_2026_1_data.py
+++ b/examples/FinRL_StockTrading_2026_1_data.py
@@ -6,14 +6,21 @@
Introduce how to use FinRL to fetch and process data that we need for ML/RL trading.
"""
+from __future__ import annotations
+
import itertools
import pandas as pd
import yfinance as yf
from finrl import config_tickers
-from finrl.config import INDICATORS, TRAIN_START_DATE, TRAIN_END_DATE, TRADE_START_DATE, TRADE_END_DATE
-from finrl.meta.preprocessor.preprocessors import FeatureEngineer, data_split
+from finrl.config import INDICATORS
+from finrl.config import TRADE_END_DATE
+from finrl.config import TRADE_START_DATE
+from finrl.config import TRAIN_END_DATE
+from finrl.config import TRAIN_START_DATE
+from finrl.meta.preprocessor.preprocessors import data_split
+from finrl.meta.preprocessor.preprocessors import FeatureEngineer
from finrl.meta.preprocessor.yahoodownloader import YahooDownloader
# %% Part 1. Fetch data - Single ticker
diff --git a/examples/FinRL_StockTrading_2026_2_train.py b/examples/FinRL_StockTrading_2026_2_train.py
index 457a2b3223..c03d7a5401 100644
--- a/examples/FinRL_StockTrading_2026_2_train.py
+++ b/examples/FinRL_StockTrading_2026_2_train.py
@@ -7,11 +7,15 @@
Introduce how to use FinRL to make data into the gym form environment, and train DRL agents on it.
"""
+from __future__ import annotations
+
import pandas as pd
from stable_baselines3.common.logger import configure
from finrl.agents.stablebaselines3.models import DRLAgent
-from finrl.config import INDICATORS, TRAINED_MODEL_DIR, RESULTS_DIR
+from finrl.config import INDICATORS
+from finrl.config import RESULTS_DIR
+from finrl.config import TRAINED_MODEL_DIR
from finrl.main import check_and_make_directories
from finrl.meta.env_stock_trading.env_stocktrading import StockTradingEnv
diff --git a/examples/FinRL_StockTrading_2026_3_Backtest.py b/examples/FinRL_StockTrading_2026_3_Backtest.py
index 63a64bda68..f3e8b25c34 100644
--- a/examples/FinRL_StockTrading_2026_3_Backtest.py
+++ b/examples/FinRL_StockTrading_2026_3_Backtest.py
@@ -8,7 +8,10 @@
Mean Variance Optimization and DJIA index.
"""
+from __future__ import annotations
+
import matplotlib
+
matplotlib.use("Agg")
import matplotlib.pyplot as plt
import numpy as np
diff --git a/finrl/config_tickers.py b/finrl/config_tickers.py
index f90efb8b68..60fd2514ed 100644
--- a/finrl/config_tickers.py
+++ b/finrl/config_tickers.py
@@ -140,7 +140,7 @@
"TRI",
"WBD",
"WMT",
- "ZS"
+ "ZS",
]
# SP 500 constituents at 2019
diff --git a/finrl/meta/env_stock_trading/env_stocktrading.py b/finrl/meta/env_stock_trading/env_stocktrading.py
index 2fb3a69c84..6516e29d24 100644
--- a/finrl/meta/env_stock_trading/env_stocktrading.py
+++ b/finrl/meta/env_stock_trading/env_stocktrading.py
@@ -28,6 +28,9 @@ class StockTradingEnv(gym.Env):
sell_cost_pct (float, array): Cost for selling shares, each index corresponds to each asset
turbulence_threshold (float): Maximum turbulence allowed in market for purchases to occur. If exceeded, positions are liquidated
print_verbosity(int): When iterating (step), how often to print stats about state of env
+ allow_short_selling (bool): If False, the agent cannot take short
+ positions. The action space is restricted to [0, 1] and negative
+ actions are clipped. Default is True for backward compatibility.
"""
metadata = {"render.modes": ["human"]}
@@ -55,6 +58,7 @@ def __init__(
model_name="",
mode="",
iteration="",
+ allow_short_selling: bool = True,
):
self.day = day
self.df = df
@@ -68,7 +72,10 @@ def __init__(
self.state_space = state_space
self.action_space = action_space
self.tech_indicator_list = tech_indicator_list
- self.action_space = spaces.Box(low=-1, high=1, shape=(self.action_space,))
+ if self.allow_short_selling:
+ self.action_space = spaces.Box(low=-1, high=1, shape=(self.action_space,))
+ else:
+ self.action_space = spaces.Box(low=0, high=1, shape=(self.action_space,))
self.observation_space = spaces.Box(
low=-np.inf, high=np.inf, shape=(self.state_space,)
)
@@ -83,6 +90,7 @@ def __init__(
self.model_name = model_name
self.mode = mode
self.iteration = iteration
+ self.allow_short_selling = allow_short_selling
# initalize state
self.state = self._initiate_state()
@@ -315,6 +323,9 @@ def step(self, actions):
actions = actions.astype(
int
) # convert into integer because we can't by fraction of shares
+ # Prevent short selling: clip negative actions to 0
+ if not self.allow_short_selling:
+ actions = np.where(actions < 0, 0, actions)
if self.turbulence_threshold is not None:
if self.turbulence >= self.turbulence_threshold:
actions = np.array([-self.hmax] * self.stock_dim)