rhoa 0.2.1

A pandas DataFrame extension for technical analysis.

Rhoa is a Python package providing pandas DataFrame extension accessors for technical analysis and machine learning in financial markets.

---

🎯 Features

• 📊 13 Technical Indicators - Professional-grade indicators accessible via pandas Series extension

  • Moving averages (SMA, EWMA)
  • Momentum oscillators (RSI, MACD, Stochastic, Williams %R, CCI)
  • Volatility indicators (ATR, Bollinger Bands, EWMSTD)
  • Trend indicators (ADX, Parabolic SAR)

• 🤖 ML Target Generation - Intelligent binary target generation for machine learning

  • Auto mode with Pareto optimization
  • Manual mode with elbow method
  • 8 different target calculation methods
  • Optimal threshold detection

• 📈 Visualization - Professional charting for prediction analysis

  • Price charts with buy signals
  • Confusion matrices
  • Color-coded performance markers
  • Publication-quality output

• 🔗 Pandas Integration - Seamless integration with pandas workflows

  • Native DataFrame/Series accessors
  • Method chaining support
  • Type hints throughout

🚀 Quick Start

Installation

pip install rhoa

Basic Usage

import pandas as pd
import rhoa

# Load your price data
df = rhoa.read_csv('stock_prices.csv')

# Calculate technical indicators using Series accessor
df['SMA_20'] = df['Close'].rhoa.indicators.sma(window_size=20)
df['RSI_14'] = df['Close'].rhoa.indicators.rsi(window_size=14)

# Calculate MACD
macd_data = df['Close'].rhoa.indicators.macd()
df['MACD'] = macd_data['macd']
df['Signal'] = macd_data['signal']

# Calculate Bollinger Bands
bb = df['Close'].rhoa.indicators.bollinger_bands(window_size=20, num_std=2.0)
df['BB_Upper'] = bb['upper_band']
df['BB_Lower'] = bb['lower_band']

Generate ML Targets

from rhoa.targets import generate_target_combinations

# Generate optimized binary targets with Pareto optimization
targets, metadata = generate_target_combinations(
    df,
    mode='auto',
    target_class_balance=0.5  # Aim for 50% positive instances
)

# Check what parameters were found optimal
print(metadata['method_7'])
# {'period': 6, 'threshold': 4.0, 'instances': 249, 'pct_of_max': 8.9}

# Combine with features for ML pipeline
ml_data = pd.concat([df, targets], axis=1)

Visualize Predictions

# After training your model and getting predictions
predictions = model.predict(X_test)

# Plot with confusion matrix
fig = df.rhoa.plots.signal(
    y_pred=predictions,
    y_true=y_test,
    date_col='Date',
    price_col='Close'
)

📚 Documentation

Full documentation is available at https://nainajnaho.github.io/Rhoa/

• Installation Guide
• Quick Start Tutorial
• User Guide
• API Reference
• Examples
• FAQ

💡 Examples

Example 1: Find Overbought Conditions

import pandas as pd
import rhoa

df = rhoa.read_csv('prices.csv')

# Calculate RSI
rsi = df['Close'].rhoa.indicators.rsi(window_size=14)

# Find overbought periods (RSI > 70)
overbought = df[rsi > 70]
print(f"Found {len(overbought)} overbought periods")

Example 2: Detect MACD Crossovers

# Calculate MACD
macd_data = df['Close'].rhoa.indicators.macd()
macd = macd_data['macd']
signal = macd_data['signal']

# Find bullish crossovers
bullish = (macd > signal) & (macd.shift(1) <= signal.shift(1))
print(f"Bullish crossovers: {bullish.sum()}")

Example 3: Complete ML Pipeline

from rhoa.targets import generate_target_combinations
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split

# Generate features
df['SMA_20'] = df['Close'].rhoa.indicators.sma(20)
df['RSI_14'] = df['Close'].rhoa.indicators.rsi(14)
df['ATR_14'] = df['Close'].rhoa.indicators.atr(df['High'], df['Low'], 14)

# Generate targets
targets, meta = generate_target_combinations(df, mode='auto')

# Prepare data
X = df[['SMA_20', 'RSI_14', 'ATR_14']].dropna()
y = targets['Target_7'].loc[X.index]

# Train model
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
model = RandomForestClassifier(n_estimators=100, random_state=42)
model.fit(X_train, y_train)

print(f"Accuracy: {model.score(X_test, y_test):.2%}")

🛠️ Development Status

Rhoa is currently in pre-alpha development (v0.1.7). The API may change between versions.

Roadmap

• Core technical indicators (In progress...)
• ML target generation (In progress...)
• Visualization tools (In progress...)
• Strategy framework
• Advanced preprocessing utilities
• Backtesting engine

📋 Requirements

• Python >= 3.9
• pandas >= 1.3
• numpy >= 1.21

Optional dependencies for specific features:

• kneed - For elbow method in target generation
• paretoset - For Pareto optimization
• scikit-learn - For confusion matrices
• matplotlib - For visualization
• seaborn - For enhanced plotting

📄 License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.

🔗 Links

• Documentation: https://nainajnaho.github.io/Rhoa/
• PyPI: https://pypi.org/project/rhoa/
• Source Code: https://github.com/nainajnahO/Rhoa
• Issue Tracker: https://github.com/nainajnahO/Rhoa/issues
• Changelog: CHANGELOG.md

📊 Project Status

---