TRACES

Time-series Relationship Analysis with Comprehensive Evaluation Suite

A specialized Python framework for multi-method time series correlation analysis with automatic method optimization.

Overview

TRACES automatically analyzes relationships between time series data using multiple correlation methods (Pearson, Spearman, Kendall, CCF) and determines the most appropriate method(s) for each pair.

Key Features

• Automatic detection of relationship types (linear, non-linear, lagged, complex)
• Handles hierarchical data structures (parent-child relationships)
• Comprehensive visualization suite
• Multi-method correlation analysis
• Statistical significance testing

Requirements

• Python 3.8+
• pandas
• numpy
• scipy
• matplotlib
• seaborn

Input Data Format

• Excel file (.xlsx)
• First row: Column headers (series names)
• First column: Time intervals
• Additional columns: Time series data

Quick Start

# Example usage
import pandas as pd
from traces import analyze_full_dataset

# Load your data
df = pd.read_excel('your_data.xlsx')

# Run analysis
results = analyze_full_dataset(df, valid_pairs)

# View results
print(results.sort_values('Abs_Max_Corr', ascending=False))