astrodata 2.9.3

A package for managing astronomical data through a uniform interface

astrodata

Tests

Building & Publishing

Making astronomical data consistent and approachable

astrodata is a package for managing astronomical data through a uniform interface. It is designed to be used with the Astropy package. astrodata was created for use as part of the DRAGONS data reduction pipeline, but it is now implemented to be useful for any astronomical data reduction or analysis project.

Unlike managing files using the astropy.io.fits package alone, astrodata is designed to be extendible to any data format, and to parse, respond to, and store metadata in a consistent, intentional way. This makes it especially useful for managing data from multiple instruments, telescopes, and data generation utilities.

Note: If you are trying to reduce Gemini data, please use DRAGONS. Interaction with this package directly is primarily suited for developers, and does not come with any tools for data reduction on any specific instrument or data.

Installation

astrodata is available on the Python Package Index and can be installed using pip:

python -m pip install astrodata

Documentation

Documentation for astrodata is available on our GitHub pages site. This documentation includes a user and programmer's guide, as well as a full API reference.

Usage

The most basic usage of astrodata is to extend the astrodata.AstroData class, which includes some basic FITS file handling methods by default:

from astrodata import AstroData, astro_data_descriptor, factory, from_file


class MyData(AstroData):
    @astro_data_descriptor
    def color(self):
        # The color filter used for our image is stored in a few different
        # ways, let's unify them.
        blue_labels = {"blue", "bl", "b"}
        green_labels = {"green", "gr", "g"}
        red_labels = {"red", "re", "r"}

        header_value = self.phu.get("COLOR", None).casefold()

        if header_value in blue_labels:
            return "BLUE"

        if header_value in green_labels:
            return "GREEN"

        if header_value in red_labels:
            return "RED"

        if header_value is None:
            raise ValueError("No color found")

        # Unrecognized color
        raise ValueError(f"Did not recognize COLOR value: {header_value}")


# Now, define our instruments with nuanced, individual data formats
class MyInstrument1(MyData):
    # These use a special method to resolve the metadata and apply the correct
    # class.
    @staticmethod
    def _matches_data(source):
        return source[0].header.get("INSTRUME", "").upper() == "MYINSTRUMENT1"


class MyInstrument2(MyData):
    @staticmethod
    def _matches_data(source):
        return source[0].header.get("INSTRUME", "").upper() == "MYINSTRUMENT2"


class MyInstrument3(MyData):
    @staticmethod
    def _matches_data(source):
        return source[0].header.get("INSTRUME", "").upper() == "MYINSTRUMENT3"


for cls in [MyInstrument1, MyInstrument2, MyInstrument3]:
    factory.add_class(cls)

# my_file.fits has some color data depending on the instrument it comes from,
# but now we can access it and handle a single value.
data = from_file("README_example.fits")

# the astrodata factory has already resolved the correct class for us.
print(f"File used to create class: {data.__class__.__name__}")
if data.color() == "BLUE":
    print("I used the blue filter!")

else:
    print("I used a red or green filter!")

# Get all the info about the astrodata object.
data.info()

This will print out the filter used as extracted from the header of the FITS file. data.info() offers a more complete look at the file's data including the filename and path of the file (as it does for astropy.io.fits objects).

astrodata is designed to be extensible, so you can add your own methods to analyze and process data based on your specific needs and use cases.

For a complete example, see the Quickstart in our documentation.

Installing development dependencies

astrodata uses Poetry for build and package management. Our documentation includes an installation guide for astrodata developers

Contributing

See our contributing guidelines for information on contributing. If you're worried about contributing, or feel intimidated, please remember that your contribution is immensely appreciated---no matter how small!

License

This project is Copyright 2024 (c) and licensed under the terms of a modified BSD 3-clause license through AURA astronomy. This package is based upon the Openastronomy packaging guide which is licensed under the standard BSD 3-clause license. See the LICENSE file for more information.

Citations

To cite astrodata in your work, please see CITATION.md for complete information, including a bibtex example.

For ease of reference, the current citation to use is: Simpson et al. 2024.