punchbowl.data.punch_io#

Attributes#

`_ROOT`
`CALIBRATION_ANNOTATION`

Classes#

DefaultFormatter

A formatter that doesn't fail if a keyword is missing. Used for quicklook.

Functions#

`write_file_hash`(→ None)	Create a SHA-256 hash for a file.
`get_base_file_name`(→ str)	Determine the base file name without file type extension.
`_header_to_xml`(→ lxml.etree.Element)	Convert image header metadata into an XML Tree that can be inserted into a JPEG2000 file header.
`_generate_jp2_xmlbox`(→ glymur.jp2box.XMLBox)	Generate the JPEG2000 XML box to be inserted into the JPEG2000 file.
`write_ndcube_to_quicklook`(→ None)	Write an NDCube to a Quicklook format as a jpeg.
`write_quicklook_to_mp4`(→ None)	Write a list of input quicklook jpeg2000 files to an output mp4 animation.
`write_ndcube_to_fits`(→ None)	Write an NDCube as a FITS file.
`_pack_uncertainty`(→ numpy.ndarray)	Compress the uncertainty for writing to file.
`_unpack_uncertainty`(→ numpy.ndarray)	Uncompress the uncertainty when reading from a file.
`_update_statistics`(...)	Update image statistics in metadata before writing to file.
`load_ndcube_from_fits`(→ ndcube.NDCube)	Load an NDCube from a FITS file.
`_load_many_cubes_caller`(→ ndcube.NDCube \| str)
`load_many_cubes_iterable`(→ list[ndcube.NDCube \| str])	Load many NDCubes in parallel.
`load_many_cubes`(→ list[ndcube.NDCube \| str])	Load many NDCubes in parallel.
`check_outlier`(→ bool)	Check the input data cube for outlier flagging.

Module Contents#

punchbowl.data.punch_io._ROOT = b'.'#

punchbowl.data.punch_io.CALIBRATION_ANNOTATION = '{OBSRVTRY} - {TYPECODE}{OBSCODE} - {DATE-OBS} - exptime: {EXPTIME} s - polarizer: {POLAR} deg'#

punchbowl.data.punch_io.write_file_hash(path: str) → None[source]#: Create a SHA-256 hash for a file.

punchbowl.data.punch_io.get_base_file_name(cube: ndcube.NDCube) → str[source]#: Determine the base file name without file type extension.

class punchbowl.data.punch_io.DefaultFormatter[source]#

Bases: string.Formatter

A formatter that doesn’t fail if a keyword is missing. Used for quicklook.

get_field(field_name: str, args: Any, kwargs: Any) → str[source]#: Provide a special getter that returns the name if it fails.

punchbowl.data.punch_io._header_to_xml(header: astropy.io.fits.Header) → lxml.etree.Element[source]#

Convert image header metadata into an XML Tree that can be inserted into a JPEG2000 file header.

(Helper function adapted from SunPy)

punchbowl.data.punch_io._generate_jp2_xmlbox(header: astropy.io.fits.Header) → glymur.jp2box.XMLBox[source]#

Generate the JPEG2000 XML box to be inserted into the JPEG2000 file.

(Helper function adapted from SunPy)

punchbowl.data.punch_io.write_ndcube_to_quicklook(cube: ndcube.NDCube, filename: str, layer: int | None = None, vmin: float = 1e-15, vmax: float = 8e-12, include_meta: bool = True, annotation: str | None = None, color: bool = False, gamma: float = 1 / 2.2) → None[source]#

Write an NDCube to a Quicklook format as a jpeg.

Parameters:

cube (NDCube) – data cube to visualize
filename (str) – path to save output, must end in .jp2, .j2k, .jpeg, .jpg
layer (int | None) – if the cube is 3D, then selects cube.data[layer] for visualization
vmin (float) – the lower limit value to visualize
vmax (float) – the upper limit value to visualize
include_meta (bool) – whether to include metadata in the JPEG2000 file
annotation (str | None) – a formatted string to add to the bottom of the image as a label can access metadata by key, e.g. “typecode={TYPECODE}” would write the data’s typecode into the image
color (bool) – flag to generate RGB quicklook files, grayscale by default
gamma (float) – power law exponent used for color normalization

Return type:

None

punchbowl.data.punch_io.write_quicklook_to_mp4(files: list[str], filename: str, framerate: int = 5, resolution: int = 1024, codec: str = 'libx264', ffmpeg_cmd: str = 'ffmpeg') → None[source]#

Write a list of input quicklook jpeg2000 files to an output mp4 animation.

Parameters:

files (list[str]) – List of input files to animate
filename (str) – Output filename
framerate (int, optional) – Frame rate (default 5)
resolution (int, optional) – Output resolution (default 1024)
codec (str, optional) – Codec to use for encoding. For GPU acceleration. “h264_videotoolbox” can be used on ARM Macs, “h264_nvenc” can be used on Intel machines.
ffmpeg_cmd (str) – path to the ffmpeg executable

punchbowl.data.punch_io.write_ndcube_to_fits(cube: ndcube.NDCube, filename: str, overwrite: bool = False, write_hash: bool = True, skip_stats: bool = False, skip_wcs_conversion: bool = False, uncertainty_quantize_level: float = 16) → None[source]#: Write an NDCube as a FITS file.

punchbowl.data.punch_io._pack_uncertainty(cube: ndcube.NDCube) → numpy.ndarray[source]#: Compress the uncertainty for writing to file.

punchbowl.data.punch_io._unpack_uncertainty(uncertainty_array: numpy.ndarray, data_array: numpy.ndarray) → numpy.ndarray[source]#: Uncompress the uncertainty when reading from a file.

punchbowl.data.punch_io._update_statistics(cube: ndcube.NDCube, modify_inplace: bool = False) → punchbowl.data.meta.NormalizedMetadata[source]#: Update image statistics in metadata before writing to file.

punchbowl.data.punch_io.load_ndcube_from_fits(path: str | pathlib.Path, key: str = ' ', include_provenance: bool = True, include_uncertainty: bool = True, dtype: type = float) → ndcube.NDCube[source]#: Load an NDCube from a FITS file.

punchbowl.data.punch_io._load_many_cubes_caller(path: str | pathlib.Path, kwargs: dict, allow_errors: bool) → ndcube.NDCube | str[source]#

punchbowl.data.punch_io.load_many_cubes_iterable(paths: list[str | pathlib.Path], n_workers: int | None = None, allow_errors: bool = False, **kwargs: dict) → list[ndcube.NDCube | str][source]#

Load many NDCubes in parallel.

Does not fork so as to be Prefect-compatible.

When used as an iterator, cubes are yielded as they are loaded, allowing e.g. progress messages to be printed

Parameters:

paths – File paths to load.
n_workers – Number of parallel workers to use. A large number may overwhelm the main thread (which has to receive each loaded cube), limiting the speed benefits of using many workers.
allow_errors – If True, if an exception is raised when loading a cube, the traceback is yielded rather than an NDCube. If False, exceptions are raised in the normal way.
kwargs – Extra args are passed to load_NDCube_from_fits

punchbowl.data.punch_io.load_many_cubes(paths: list[str | pathlib.Path], n_workers: int | None = None, allow_errors: bool = False, **kwargs: dict) → list[ndcube.NDCube | str][source]#

Load many NDCubes in parallel.

Does not fork so as to be Prefect-compatible.

Parameters:

paths – File paths to load.
n_workers – Number of parallel workers to use. A large number may overwhelm the main thread (which has to receive each loaded cube), limiting the speed benefits of using many workers.
allow_errors – If True, if an exception is raised when loading a cube, the traceback is yielded rather than an NDCube. If False, exceptions are raised in the normal way.
kwargs – Extra args are passed to load_NDCube_from_fits

Return type:

A list of NDCubes (or traceback strings)

punchbowl.data.punch_io.check_outlier(cube: ndcube.NDCube) → bool[source]#: Check the input data cube for outlier flagging.