Virgo_Stocks

Virgo Stock is a package for financial technical analysis. This package is develop with the mind of automatic analysis.

There are four major components in this package: Data Source, Stock, Indicator, and Strategy.

Data Source

This package relies on external source to provide market data. The DataSourceInterface acts as a layer between the data source and the Stock class. The DataSourceInterface is intended to provide a unified interface for accessing market data from different data source. A sub-class should be implemented for each data source.

Data Source Interface

Each sub-class should implement two abstract methods:

• get_daily_series(), to get the daily series data.
• get_intraday_series(), to get the intra-day data.

The data source interface provides a get_stock() method to obtain a Stock instance.

Alpha Vantage

Alpha Vantage provides web API for stock data and more.

This package implemented an AlphaVantage data source as a subclass of DataSourceInterface. The implementation here includes:

1. The AlphaVantageAPI class as a simple python API for accessing the AlphaVantage data. See more details.
2. An option to cache the data to reduce the outgoing API requests.

See also: https://www.alphavantage.co/

Stock

The Stock component is designed to manipulate and transform stock/equity data.

In this package, a Stock object represents a stock (or security). Stock objects provides methods for retrieving and manipulating series data of a particular stock. Two parameters are required to initialize a Stock object:

• The symbol of the stock as a string.
• A DataSource implemented the DataSourceInterface for retrieving the data.

Once initialized, the stock series data are retrieved as a pandas DataFrame. The data frame will have at least 5 columns: open, high, low, close and volume. The first row of the data frame stores the latest data.

The Stock class provides methods to obtain a DataSeries instance.

• daily_series() returns daily data.
• intraday_series() returns intra-day data.
• weekly_series() returns weekly data.
• monthly_series() returns monthly data.

DataSeries is a sub-class of pandas DataFrame. It provides an indicator() method to obtain technical indicators.

Indicator

Technical Indicator is the basic component of technical analysis.

In this package, each indicator is defined as a class. In addition to the calculated time series data, each indicator class also class-specific provides aggregated information (e.g. Moving Average also provides Golden Crosses and Death Crosses). An indicator can be initialized with a stock series data. There are two categories of indicators:

• The ones inheriting from IndicatorSeries, which is a sub-class of pandas Series, contain single data series.
• The ones inheriting from IndicatorDataFrame, which is a sub-class of pandas DataFrame, contain multi-column data.

Strategy

The Strategy class is designed to simulate and evaluate trading strategies. Simulation is the starting point for automatic analysis. Trading strategies can be complicated. Each strategy should be implemented as a sub-class by the user.

Modules, Classes, Objects and the Relations between them

The Virgo_Stock package defines classes in the following files:

• series.py: defines TimeSeries and TimeDataFrame, which are base classes for most data types in this package.
• source.py: defines DataSourceInterface and implements the AlphaVantage data source.
• stock.py: defines Stock and DataPoint;
• indicators.py: defines Indicator as the base class and sub-classes for calculating technical indicators (e.g. moving average).
• strategy.py: defines Strategy as the base class for simulating and evaluating strategies.

2018-2020 Qiu Qin. All Right Reserved.

This is a mirrored repository, which does not contain the latest Development. This package is part of Qiu's Astrology Collection.