CoinBot: AI-Powered Trading Model

CoinBot is an AI-driven trading model designed to optimize trading decisions for individual private treasuries. Utilizing advanced algorithms and historical data analysis, CoinBot empowers traders with data-driven insights and responsive trading capabilities.

Disclaimer

I am a programmer and NOT an accredited financial expert.
For important investment decisions, consult a certified financial professional.
Do NOT rely on investment advice from anonymous sources online; always conduct thorough research.

📝 THIS SOFTWARE IS UNDER ACTIVE DEVELOPMENT AND IS SUBJECT TO CHANGE

⚠️ USE THIS SOFTWARE AT YOUR OWN RISK!

🚨 ALWAYS AUDIT THE OUTPUT!

Features

• AI-Driven Insights: Leverage AI for informed trading decisions based on market analysis.
• Real-Time Trading: Swiftly respond to market changes with CoinBot's real-time capabilities.
• Diverse Asset Support: Trade cryptocurrencies and stocks through Alpaca Markets API.
• Customizable Strategies: Tailor trading strategies to your risk tolerance and asset preferences.
• Smart Portfolio Management: Optimize portfolio allocation for returns and risk management.
• Tailored Training: Train CoinBot on your chosen assets to specialize its performance.

Getting Started

Explore CoinBot's potential. Join in shaping AI-enhanced trading for individual treasuries.

Future Developments

Stay tuned for exciting updates, including advanced tax strategies, enhanced risk management, and more.

Usage

CoinBot XOR Model

The CoinBot XOR model can be trained and utilized using the provided command-line interface (CLI) script. Follow these steps to effectively use the script:

Prerequisites

Make sure you have the necessary dependencies and a virtual environment set up. Navigate to the coinbot directory where the xor.py script is located.

Training and Prediction

1. Open a terminal window.

2. Activate the virtual environment, if not already activated, to ensure proper package dependencies are used:

   poetry shell

3. Run the script using the following command:

   python -m coinbot.xor [OPTIONS]

   Replace [OPTIONS] with the desired command-line options described below.

Command-Line Options

The script supports various options to customize the training process and model behavior. Use the following options to control the training process:

• --model_id TEXT: Specify the file path for the model. The default is 'models/coinbot-xor.h5'.

• --epochs INTEGER: Set the number of training epochs. The default is 10,000.

• --learning_rate FLOAT: Set the learning rate for the training process. The default is 0.15.

• --lambda_ FLOAT: Set the L2 regularization parameter. The default is 1e-5.

• --tolerance FLOAT: Set the tolerance for early stopping during training. The default is 1e-5.

• --help: Display a help message with the options and their descriptions.

Customizing Training

To customize the behavior of the script and train the model with specific parameters, use the command-line options. For example, to train the model with different parameters:

python -m coinbot.xor --epochs 5000 --learning_rate 0.1

Output

During the training process, the script will display log messages indicating the progress of the training. After training, the script will output the predictions from the trained model, along with metadata about the architecture and model. The trained model will be saved in the specified file path.

00:45:19 | ~/Documents/code/remote/coinbot
(.venv) git:(main | Δ) λ python -m coinbot.xor
2023-09-04 00:45:58,215 - INFO - dense.py:162 - Epoch 0, Loss: 0.9282104981979823
2023-09-04 00:45:58,241 - INFO - dense.py:162 - Epoch 1000, Loss: 0.007774923986908178
2023-09-04 00:45:58,245 - WARNING - dense.py:147 - Early stopping on epoch 1182, Loss: 0.002676929559403822
Running predictions:
[[0.00750384]
 [0.93021532]
 [0.93170549]
 [0.01551835]]
[{'type': 'Dense', 'input_dim': 2, 'output_dim': 2}, {'type': 'Tanh'}, {'type': 'Dense', 'input_dim': 2, 'output_dim': 1}, {'type': 'Tanh'}]
2023-09-04 00:45:58,247 - INFO - xor.py:89 - Saved model: models/coinbot-xor.h5

Remember that this script is tailored for the CoinBot XOR model and provides an intuitive way to fine-tune training parameters and manage the model path using command-line options.

CoinBot Simulation for Pre-Training

The simulate.py script is designed to work with a single trade pair and blocks if more than one trade pair is given, you can use the following commands to test different scenarios:

1. Basic Usage with Default Values: This command will run the script with default values for the trade pair, timeframe, and other options.

   python -m coinbot.simulate

2. Specify Trade Pair and Timeframe: You can specify the trade pair (crypto symbols) and the timeframe for sampling. For example:

   python -m coinbot.simulate --symbols ETH/USD --timeframe 1H

3. Specify Start and End Dates: You can provide specific start and end dates for sampling. For example:

   python -m coinbot.simulate --start 2023-01-01 --end 2023-08-31

4. Customize Principal and Rate: You can adjust the principal amount and annual interest rate. For example:

   python -m coinbot.simulate --principal 500.00 --rate 0.15

5. Change Frequency and Interval: You can modify the compounding frequency and time step interval. For example:

   python -m coinbot.simulate --frequency 250 --interval 2

6. Testing Error Handling: You can test how the script handles errors by providing an invalid trade pair or missing required options:

   python -m coinbot.simulate --symbols INVALIDPAIR
   python -m coinbot.simulate --symbols BTC/USD --interval -1

7. Example with All Options: A more comprehensive command using various options:

   python -m coinbot.simulate --symbols LTC/USD --timeframe 15T --start 2023-07-01 --end 2023-07-15 --principal 1000.00 --rate 0.08 --frequency 180 --interval 3

Keep in mind that the script is designed for a single trade pair, so you can experiment with different options to see how the script behaves under various scenarios. Make sure to provide valid options to avoid triggering error handling.

License

CoinBot: Automated Trading with Compound Interest
Copyright (C) 2023 Austin Berrio

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as published
by the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Affero General Public License for more details.

You should have received a copy of the GNU Affero General Public License
along with this program. If not, see <https://www.gnu.org/licenses/>.