A Python package for simulating planned and unplanned data gaps in LISA time series data. Our package is currently available on test.pypi here.

The work here builds off the work in the lisaglitch package, which provides core functionality for generating gap masks. lisa-gap extends this functionality with advanced features such as customizable smooth tapering around gap edges using Tukey windows, data segmentation with edge tapering, and proportional tapering for frequency domain analysis.

lisa-gap provides tools for generating realistic gap masks that can be applied to LISA time series data. The package supports both planned gaps (e.g., scheduled maintenance) and unplanned gaps (e.g., hardware failures) with configurable statistical distributions.

The package includes advanced features for:

• Smooth tapering around gap edges using customizable Tukey windows
• Data segmentation with edge tapering for spectral analysis
• Proportional tapering with automatic gap categorization
• Edge boundary management to prevent spectral artifacts

These features are particularly useful for frequency domain analysis where sharp discontinuities can introduce spectral artifacts.

uv is a fast Python package manager that simplifies dependency management:

# Install uv if you don't have it
curl -LsSf https://astral.sh/uv/install.sh | sh

# Clone and install the package
git clone https://github.com/ollieburke/lisa-gap.git
cd lisa-gap
uv sync

You may find that you get errors due to the Git LFS. If you do not wish to download large files, use this. This is fine if you only intend to use lisagap as a masking (and tapering) code

	GIT_LFS_SKIP_SMUDGE=1 uv pip install -e .

To use the package:

uv run python your_script.py

If you prefer using the standard Python virtual environment:

python -m venv gap_env
source gap_env/bin/activate  # On Windows use `gap_env\Scripts\activate`

Then install the package:

pip install lisa-gap

Available on Test PyPI:

pip install -i https://test.pypi.org/simple/ --extra-index-url https://pypi.org/simple lisa-gap==0.4.0

git clone https://github.com/ollieburke/lisa-gap.git
cd lisa-gap
pip install .

Recommended with uv:

git clone https://github.com/ollieburke/lisa-gap.git
cd lisa-gap
uv sync --extra dev --extra docs

Traditional method:

git clone https://github.com/ollieburke/lisa-gap.git
cd lisa-gap
pip install -e ".[dev, docs]"

With uv:

uv run pytest

Traditional:

pytest 

## Quick Start

```python
from lisaglitch import GapMaskGenerator
from lisagap import GapWindowGenerator, DataSegmentGenerator
import numpy as np

# Create time array
dt = 5.0  # seconds
duration = 86400  # 1 day in seconds
sim_t = np.arange(0, duration, dt)

# Define gap configuration
gap_definitions = {
    "planned": {
        "maintenance": {
            "rate_per_year": 12,  # 12 times per year
            "duration_hr": 2.0    # 2 hours each
        }
    },
    "unplanned": {
        "hardware_failure": {
            "rate_per_year": 4,   # 4 times per year
            "duration_hr": 0.5    # 30 minutes each
        }
    }
}

# Create gap mask generator
gap_gen = GapMaskGenerator(
    sim_t=sim_t,
    gap_definitions=gap_definitions,
    treat_as_nan=False
)

# Create window generator for advanced features
window = GapWindowGenerator(gap_gen)

# Generate gap mask with proportional tapering
gap_mask = GapWindowGenerator.apply_proportional_tapering(
    window.generate_mask(),
    dt=dt,
    short_taper_fraction=0.25,   # 25% each side for short gaps
    medium_taper_fraction=0.05,  # 5% each side for medium gaps
    long_taper_fraction=0.02     # 2% each side for long gaps
)

# Generate data stream
data = np.sin(2*np.pi * 0.01 * sim_t) + 0.1 * np.random.randn(len(sim_t))

# Option 1: Apply gaps directly
data_w_gaps = data * gap_mask

# Option 2: Use segmentation for independent analysis
segmenter = DataSegmentGenerator(
    mask=gap_mask,
    data=data,
    dt=dt
)

# Get segments with edge tapering for frequency analysis
segments = segmenter.get_time_segments(
    apply_window=True,
    left_edge_taper=1000,   # Taper first segment left edge
    right_edge_taper=1000   # Taper last segment right edge
)

# Analyze frequency content of each segment
freq_info = segmenter.get_freq_info_from_segments()

gap_definitions = { "planned": { "maintenance": { "rate_per_year": 12, # 12 times per year "duration_hr": 2.0 # 2 hours each } }, "unplanned": { "hardware_failure": { "rate_per_year": 4, # 4 times per year "duration_hr": 0.5 # 30 minutes each } } }

gap_gen = GapMaskGenerator( sim_t=sim_t, gap_definitions=gap_definitions, treat_as_nan=True )

window = GapWindowGenerator(gap_gen)

gap_mask = window.generate_mask()

data = np.sin(2*np.pi * sim_t) # Generate fake data

data_w_gaps = data * gap_mask

## Documentation

Full documentation is available at [Read the Docs](https://lisagap.readthedocs.io/).

### Tutorial Notebook

See the included **`docs/source/gap_notebook.ipynb`** for a comprehensive tutorial that covers:

- Setting up realistic gap configurations for LISA
- Generating gap masks with planned and unplanned gaps  
- Saving and loading gap configurations to/from HDF5 files. Gap metadata can be saved to .json files
- **Customizable smooth tapering** for frequency domain analysis
- **Proportional tapering** with automatic gap categorization
- **Data segmentation** with edge tapering for spectral analysis
- **Advanced windowing** techniques for boundary artifact prevention
- Quality flag generation and gap analysis

The tutorial notebook can be viewed within the documentation.

## Advanced Features

### Data Segmentation with Edge Tapering

Split data into continuous segments and apply edge tapering to prevent spectral artifacts:

```python
from lisagap import DataSegmentGenerator

# Create segmenter
segmenter = DataSegmentGenerator(mask=gap_mask, data=your_data, dt=dt)

# Get segments with edge tapering for frequency analysis
segments = segmenter.get_time_segments(
    apply_window=True,
    left_edge_taper=1000,   # Smooth left edge of first segment
    right_edge_taper=1500   # Smooth right edge of last segment  
)

# Analyze frequency content
freq_info = segmenter.get_freq_info_from_segments()

Automatically categorize gaps and apply proportional tapering:

# Apply smart tapering based on gap duration
tapered_mask = GapWindowGenerator.apply_proportional_tapering(
    gap_mask,
    dt=dt,
    short_taper_fraction=0.25,   # 25% of gap length for short gaps
    medium_taper_fraction=0.05,  # 5% of gap length for medium gaps
    long_taper_fraction=0.02,    # 2% of gap length for long gaps
    short_gap_threshold_minutes=30,   # <30 min = short gap
    long_gap_threshold_hours=2        # >2 hours = long gap
)

Flexible Taper Control: Users have complete freedom to choose their own tapering strategy for each gap type:

from lisaglitch import GapMaskGenerator
from lisagap import GapWindowGenerator
import numpy as np

# Set up gap configuration
gap_definitions = {
    "planned": {
        "antenna repointing": {"rate_per_year": 26, "duration_hr": 3.3},
        "TM stray potential": {"rate_per_year": 2, "duration_hr": 24}
    },
    "unplanned": {
        "platform safe mode": {"rate_per_year": 3, "duration_hr": 60},
        "QPD loss micrometeoroid": {"rate_per_year": 5, "duration_hr": 24}
    }
}

# Generate gap mask
gap_gen = GapMaskGenerator(sim_t, gap_definitions)
window = GapWindowGenerator(gap_gen)
gap_mask = window.generate_mask()

# Define custom tapering per gap type (lobe lengths in hours)
taper_definitions = {
    "planned": {
        "antenna repointing": {"lobe_lengths_hr": 5.0},   # Long taper for repointing
        "TM stray potential": {"lobe_lengths_hr": 0.5}    # Short taper for TM events
    },
    "unplanned": {
        "platform safe mode": {"lobe_lengths_hr": 1.0},  # Medium taper for safe mode
        "QPD loss micrometeoroid": {"lobe_lengths_hr": 2.0}  # Custom taper for QPD loss
    }
}

# Apply smooth Tukey window tapering
smoothed_mask = window.apply_smooth_taper_to_mask(
    gap_mask, 
    taper_definitions
)

This flexibility allows users to optimize tapering strategies for different gap types based on their specific analysis requirements, whether working in time or frequency domains.

• Generate realistic gap patterns for LISA time series
• Support for both planned and unplanned gaps
• Configurable gap rates and durations
• Flexible smooth tapering with user-defined Tukey windows per gap type
• Proportional tapering with automatic gap categorization
• Data segmentation with edge tapering for spectral analysis
• Advanced boundary management to prevent frequency artifacts
• Complete freedom to customize taper lengths for different gap categories
• Built on top of the robust lisaglitch package for core gap generation
• Save/load gap configurations to/from HDF5 files
• Quality flag generation and gap analysis tools
• Comprehensive documentation and tutorial notebooks

• Python ≥ 3.10
• numpy
• scipy
• h5py
• lisaglitch
• lisaconstants

MIT License

If you use this package in your research, please cite:

@software{lisagap,
  author = {Burke, Ollie and Castelli, Eleonora},
  title = {lisa-gap: A tool for simulating data gaps in LISA time series},
  url = {https://github.com/ollieburke/lisa-gap},
  version = {0.1.0},
  year = {2025}
}

Contributions are welcome! Please see our Contributing Guide for detailed information on how to contribute to the project.

This gap generation tool is suitable for LISA data processing pipelines including L01, SIM, L2D, and L2A data products.