babamul

Python client for consuming ZTF/LSST astronomical transient alerts from Babamul Kafka streams.

Installation

uv add babamul

or

pip install babamul

Quick Start

Set your credentials via environment variables or a local .env, then start consuming alerts:

export BABAMUL_KAFKA_USERNAME="your_username"
export BABAMUL_KAFKA_PASSWORD="your_password"
export BABAMUL_API_TOKEN="your_api_token"

from babamul import AlertConsumer

# Iterate over alerts (credentials loaded from env vars or .env)
with AlertConsumer(topics=["babamul.ztf.lsst-match.hosted"]) as consumer:
    for alert in consumer:
        print(
            f"{alert.objectId}: RA={alert.candidate.ra:.4f}, "
            f"Dec={alert.candidate.dec:.4f}"
        )
        break

Configuration

Avoid hardcoding credentials in code. Prefer environment variables or a local .env file that is not committed to version control. Passing username/password in code should be limited to one-off REPL usage.

Environmental Variables

export BABAMUL_KAFKA_USERNAME="your_username"
export BABAMUL_KAFKA_PASSWORD="your_password"
export BABAMUL_API_TOKEN="your_api_token"
export BABAMUL_KAFKA_SERVER="kaboom.caltech.edu:9093"  # Optional

or define these in a .env file kept out of version control and read with the python-dotenv package.

Constructor Options

Use the constructor for runtime options like offsets and timeouts.

from babamul import AlertConsumer

with AlertConsumer(
    topics=["babamul.ztf.lsst-match.hosted"],  # Topic(s) to subscribe to
    offset="earliest",  # "latest" or "earliest"
    timeout=30.0,  # Seconds to wait for messages (None = forever)
    group_id="my-consumer-group",  # Optional, auto-generated if not set
) as consumer:
    # Consume alerts here
    for alert in consumer:
        print(
            f"{alert.objectId}: RA={alert.candidate.ra:.4f}, "
            f"Dec={alert.candidate.dec:.4f}"
        )

Working with Alerts

Alert Properties

from babamul import AlertConsumer

with AlertConsumer(topics=["babamul.ztf.lsst-match.hosted"]) as consumer:
    for alert in consumer:
        # Basic info
        print(f"  Object ID: {alert.objectId}")
        print(f"  Candidate ID: {alert.candid}")
        print(f"  Position: RA={alert.candidate.ra:.6f}, Dec={alert.candidate.dec:.6f}")
        print(f"  Time: {alert.candidate.datetime.isoformat()} (JD={alert.candidate.jd:.5f})")
        print(f"  Magnitude: {alert.candidate.magpsf:.2f}±{alert.candidate.sigmapsf:.2f}")

Photometry / Light Curves

from babamul import AlertConsumer

with AlertConsumer(topics=["babamul.ztf.lsst-match.hosted"]) as consumer:
    for alert in consumer:
        for phot in alert.get_photometry(): # Full light curve
            if phot.magpsf is not None:
                print(f"  JD {phot.jd:.5f}: {phot.magpsf:.2f} mag ({phot.band})")
            else:
                print(f"  JD {phot.jd:.5f}: non-detection, limit={phot.diffmaglim:.2f} ({phot.band})")

Cutouts

from babamul import AlertConsumer

with AlertConsumer(topics=["babamul.ztf.lsst-match.hosted"]) as consumer:
    for alert in consumer:
        alert.show_cutouts()  # Displays science, template, and difference images

Context Manager

For proper resource cleanup:

from babamul import AlertConsumer

with AlertConsumer(topics=["babamul.ztf.lsst-match.hosted"]) as consumer:
    for i, alert in enumerate(consumer):
        # process alerts
        if i >= 100:
            break
# Consumer is automatically closed

Error Handling

from babamul import AlertConsumer, AuthenticationError, BabamulConnectionError

try:
    with AlertConsumer(
        topics=["babamul.ztf.lsst-match.hosted"],
    ) as consumer:
        for alert in consumer:
            # process alerts
            pass
except AuthenticationError:
    print("Invalid credentials")
except BabamulConnectionError:
    print("Cannot connect to Kafka server")

Available Topics

Babamul provides several topic categories based on survey and classification:

LSST Topics

LSST-only (no ZTF counterpart):

Topic	Description
babamul.lsst.no-ztf-match.stellar	Alerts classified as stellar
babamul.lsst.no-ztf-match.hosted	Alerts with a host galaxy
babamul.lsst.no-ztf-match.hostless	Alerts without a host galaxy
babamul.lsst.no-ztf-match.unknown	Unclassified alerts

LSST with ZTF match:

Topic	Description
babamul.lsst.ztf-match.stellar	Alerts classified as stellar
babamul.lsst.ztf-match.hosted	Alerts with a host galaxy
babamul.lsst.ztf-match.hostless	Alerts without a host galaxy
babamul.lsst.ztf-match.unknown	Unclassified alerts

ZTF Topics

ZTF-only (no LSST counterpart):

Topic	Description
babamul.ztf.no-lsst-match.stellar	Alerts classified as stellar
babamul.ztf.no-lsst-match.hosted	Alerts with a host galaxy
babamul.ztf.no-lsst-match.hostless	Alerts without a host galaxy
babamul.ztf.no-lsst-match.unknown	Unclassified alerts

ZTF with LSST match:

Topic	Description
babamul.ztf.lsst-match.stellar	Alerts classified as stellar
babamul.ztf.lsst-match.hosted	Alerts with a host galaxy
babamul.ztf.lsst-match.hostless	Alerts without a host galaxy
babamul.ztf.lsst-match.unknown	Unclassified alerts

Wildcard Subscriptions

You can use wildcards to subscribe to multiple topics:

from babamul import AlertConsumer
# All LSST topics
with AlertConsumer(topics=["babamul.lsst.*"], ...) as consumer:
    pass

# All ZTF topics with LSST matches
with AlertConsumer(topics=["babamul.ztf.lsst-match.*"], ...) as consumer:
    pass

# All hosted alerts from both surveys
with AlertConsumer(topics=["babamul.*.*.hosted"], ...) as consumer:
    pass

Development Setup

For development and testing, use a .env file to manage your credentials:

# 1. Copy the example file
cp tests/.env.example tests/.env

# 2. Edit tests/.env with your credentials
#    Get credentials at: https://babamul.caltech.edu/signup
nano tests/.env

# 3. Load automatically when running tests
# The .env file is gitignored and will not be committed

Your tests/.env file should look like:

BABAMUL_KAFKA_USERNAME=your_username
BABAMUL_KAFKA_PASSWORD=your_password
BABAMUL_API_TOKEN=your_api_token

Most examples and tests will automatically load credentials from .env using python-dotenv.

Acknowledgments

The Babamul alerts broker and BOOM software infrastructure (du Laz et al. 2026) is co-developed by the California Institute of Technology and the University of Minnesota. This work acknowledges support from the National Science Foundation through AST Award No. 2432476 (PI Kasliwal; co-PI Coughlin) and leverages experience from the Zwicky Transient Facility (co-PIs Graham and Kasliwal).