Loading Data¶

The OVRO-LWA Portal provides the open_dataset() function for loading data from various sources.

Basic Usage¶

import ovro_lwa_portal as ovro

# Load from local zarr store
ds = ovro.open_dataset("/path/to/observation.zarr")

Once loaded, the dataset has a radport accessor providing over 40 analysis methods.

Supported Data Sources¶

Local Files¶

ds = ovro.open_dataset("/path/to/observation.zarr")

Remote URLs¶

S3 (Amazon Web Services)¶

ds = ovro.open_dataset("s3://bucket-name/observation.zarr")

HTTPS¶

ds = ovro.open_dataset("https://example.com/data.zarr")

DOI-based Access¶

ds = ovro.open_dataset("doi:10.5281/zenodo.1234567")

Chunking¶

Control memory usage with chunking:

# Automatic chunking (default)
ds = ovro.open_dataset("path/to/data.zarr", chunks="auto")

# Custom chunking
ds = ovro.open_dataset(
    "path/to/data.zarr",
    chunks={"time": 100, "frequency": 50}
)

# No chunking (caution: loads all data into memory)
ds = ovro.open_dataset("path/to/data.zarr", chunks=None)

Dataset Structure¶

OVRO-LWA datasets typically contain:

Dimensions: time, frequency, l, m (and optionally beam)
Coordinates: Time stamps, frequency channels, sky coordinates
Data Variables: intensity, SKY, BEAM, etc.
Attributes: Observation metadata

Accessing the radport Accessor¶

After loading, use the .radport accessor for analysis:

ds = ovro.open_dataset("path/to/data.zarr")

# The radport accessor is now available
ds.radport.plot()
ds.radport.dynamic_spectrum()
ds.radport.find_peaks()

See the API Reference for all available methods.