penhsuanwang / yfinance-stock-analyzer-lstm Goto Github PK

Description

The current implementation of data_fetcher.py includes start_date and end_date in the constructors of BaseLoader, YahooFinanceDataLoader, and CsvLoader classes. This design limits the flexibility of the module, as it requires the date range to be set at the time of object creation. The proposed refactor aims to enhance the flexibility by allowing the date range to be specified at the time of data retrieval.

Proposed Changes

1. Remove Date Initialization from BaseLoader: The start_date and end_date initialization should be removed from the BaseLoader class.

2. Update YahooFinanceDataLoader and CsvLoader Classes: These classes should no longer receive date parameters during initialization. Instead, the date parameters should be included in the methods responsible for data retrieval.

3. Modify Data Retrieval Methods: The methods in YahooFinanceDataLoader and CsvLoader for fetching data should be updated to accept start_date and end_date as parameters.

Expected Benefits

• Increased Flexibility: Allowing dynamic specification of the date range at the time of data retrieval will make the module more versatile.
• Better Adaptability: This change will enable the module to better adapt to different use cases where the date range is not known in advance.
• Code Simplification: Simplifying the initialization of the loaders can lead to more readable and maintainable code.

Additional Notes

This is a significant change and might require careful consideration of backward compatibility and potential impacts on existing codebases using this module.

Issue Description

We need to enhance our existing stock market analysis project by integrating a module that can reliably identify and analyze various candlestick chart patterns. These patterns are crucial for understanding market sentiments and can greatly aid in making informed trading decisions.

Specific Patterns to Identify

The module should be capable of recognizing the following candlestick patterns:

Bullish Candlestick Patterns

These patterns often indicate a potential reversal from a downtrend to an upward trend:

1. Hammer: Short body with a long lower wick, showing strong buying pressure.
2. Inverse Hammer: Long upper wick, indicating buyers gaining market control.
3. Bullish Engulfing: A small red body completely engulfed by a larger green candle.
4. Piercing Line: A two-stick pattern showing strong buying pressure.
5. Morning Star: A three-stick pattern signaling a weakening selling pressure and a potential bull market.
6. Three White Soldiers: Three long green candles, a strong bullish signal.

Bearish Candlestick Patterns

These patterns typically suggest a reversal from an uptrend to a downward trend:

1. Hanging Man: Occurs at the end of an uptrend, indicating potential loss of bull control.
2. Shooting Star: Formed in an uptrend, suggesting a possible market downturn.
3. Bearish Engulfing: A small green body engulfed by a large red body.
4. Evening Star: A three-candlestick pattern signaling an uptrend reversal.
5. Three Black Crows: Three long red candles, indicating a bearish downtrend.
6. Dark Cloud Cover: A pattern signaling a bearish reversal.

Continuation Candlestick Patterns

Patterns indicating market rest or indecision:

1. Doji: Struggle between buyers and sellers with no clear winner.
2. Spinning Top: Represents market indecision.
3. Falling Three Methods: A bearish continuation pattern.
4. Rising Three Methods: A bullish continuation pattern.

Objective

The objective of this module is to provide clear and accurate identification of these patterns to enhance our market trend analysis and trading strategy formulation.

Potential Challenges

• Ensuring high accuracy in pattern recognition.
• Integrating this module seamlessly with the existing system.
• Balancing computational efficiency with the complexity of pattern analysis.

Additional Notes

This feature will be a significant step towards making our tool more robust for technical analysis. It's essential that the implementation is not only accurate but also efficient to maintain the overall performance of the system.