imperfekt 0.2.3

A framework to analyze imperfections (missingness, noise) in time-series datasets.

Imperfekt - Understanding Data Imperfections in Time-Series

A comprehensive analysis toolkit for studying "imperfect" data patterns in time-series datasets. Imperfection refers to missingness, noise, and other data quality issues that can be indicated using a binary mask.

Overview

This library provides tools to analyze data quality issues in time-series data, including:

• Intravariable analysis of imperfection patterns for individual variables
• Intervariable analysis of co-occurring imperfections across multiple parameters
• Feature generation based on missingness patterns for downstream ML tasks

Installation

Install the library using pip:

pip install imperfekt

For additional features:

# With visualization export support
pip install imperfekt[viz]

# With statistical tests
pip install imperfekt[stats]

# Full installation
pip install imperfekt[full]

Quick Start

import polars as pl
from imperfekt import Imperfekt, FeatureGenerator

# Load your time-series data
df = pl.read_parquet("your_data.parquet")

# Run full imperfection analysis
analyzer = Imperfekt(
    df=df,
    id_col="id",           # Unique identifier column
    clock_col="clock",     # Timestamp column
    cols=["var1", "var2"], # Variables to analyze
    save_path="./results"
)
results = analyzer.run()

# Or generate missingness-aware features for ML
fg = FeatureGenerator(
    df=df,
    id_col="id",
    clock_col="clock",
    variable_cols=["var1", "var2"]
)
features_df = fg.add_binary_masks().add_temporal_features().df

Library Structure

imperfekt/
├── analysis/
│   ├── preliminary/     # Basic data exploration
│   ├── intravariable/      # Single variable analysis
│   ├── intervariable/    # Multi-variable patterns
│   └── utils/           # Shared utilities
├── features/            # Feature engineering
│   ├── core.py          # FeatureGenerator class
│   ├── temporal.py      # Time-based features
│   └── interaction.py   # Variable interactions
└── config/              # Default settings

Data Format

The library expects time-series data with the following structure:

Column	Description
id	Unique identifier for each time-series (e.g., patient, sensor)
clock	Timestamp for each observation
var1, var2, ...	Variables to analyze

Key Dependencies

• polars: High-performance data processing
• plotly: Interactive visualizations
• scipy: Statistical computations

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.