Added HiPS and HiPS3D dask array classes

reproject/hips/dask_array.py

@@ -0,0 +1,278 @@
+import os
+import struct
+import urllib
+import uuid
+import functools
+
+import numpy as np
+from astropy import units as u
+from astropy.io import fits
+from astropy.wcs import WCS
+from astropy_healpix import HEALPix, level_to_nside
+from dask import array as da
+from astropy.utils.data import download_file
+
+from .utils import is_url, load_properties, tile_filename, tile_filename_3d
+
+__all__ = ['hips_as_dask', 'hips3d_as_dask']
+
+
+class HiPSArray:
+
+    def __init__(self, directory_or_url, level=None):
+
+        self._directory_or_url = directory_or_url
+
+        self._is_url = is_url(directory_or_url)
+
+        self._properties = load_properties(directory_or_url)
+
+        self._tile_width = int(self._properties["hips_tile_width"])
+        self._order = int(self._properties["hips_order"])
+        self._level = self._order if level is None else level
+
+        self._tile_format = self._properties["hips_tile_format"]
+
+        self._nside = level_to_nside(self._level)
+        self._hp = HEALPix(nside=self._nside, frame="icrs", order="nested")
+        self._cdelt = 45 / self._tile_width / 2**self._level * 2**0.5
+
+        image_size = 5 * self._nside * self._tile_width
+
+        self.wcs = WCS(naxis=2)
+        self.wcs.wcs.ctype = "RA---HPX", "DEC--HPX"
+        self.wcs.wcs.crval = 0, 0
+        self.wcs.wcs.cdelt = self._cdelt, self._cdelt
+        self.wcs.wcs.crpix = image_size / 2, image_size / 2
+        self.wcs.wcs.crota = 0, 45
+        self.wcs.wcs.set()
+
+        self.shape = (image_size, image_size)
+        self.dtype = float
+        self.ndim = 2
+
+        self.chunksize = (self._tile_width, self._tile_width)
+
+        self._nan = np.nan * np.ones(self.chunksize, dtype=self.dtype)
+
+    def __getitem__(self, item):
+
+        # For now assume item is a list of slices. Find
+
+        imid = (item[0].start + item[0].stop) // 2
+        jmid = (item[1].start + item[1].stop) // 2
+
+        # Convert pixel coordinates to HEALPix indices
+
+        index = self._hp.skycoord_to_healpix(self.wcs.pixel_to_world(imid, jmid))
+
+        if index == -1:
+            return self._nan
+
+        return self._get_tile(level=self._level, index=index)
+
+    @functools.lru_cache(maxsize=128)
+    def _get_tile(self, *, level, index):
+
+        filename_or_url = tile_filename(
+            level=self._level,
+            index=index,
+            output_directory=self._directory_or_url,
+            extension="fits",
+        )
+
+        if self._is_url:
+            try:
+                filename = download_file(filename_or_url, cache=True)
+            except urllib.error.HTTPError:
+                return self._nan
+        elif not os.path.exists(filename_or_url):
+            return self._nan
+        else:
+            filename = filename_or_url
+
+        with fits.open(filename) as hdulist:
+            hdu = hdulist[0]
+            data = hdu.data[::-1]
+
+        return data
+
+
+def freq2pix(order, freq):
+    hash_value = get_hash(freq)
+    return hash_value >> (59 - order)
+
+
+def get_hash(param_double):
+    l1 = struct.unpack(">q", struct.pack(">d", param_double))[0]
+    l2 = (l1 & 0x7FF0000000000000) >> 52
+    l2 = (l2 - 929) << 52
+    return (l1 & 0x800FFFFFFFFFFFFF) | l2
+
+
+def pix2freq(order, pix):
+    delta_order = 59 - order
+    nb_pix_in = pow2(delta_order)
+    hash_value = (pix << delta_order) + nb_pix_in // 2
+    return get_freq(hash_value)
+
+
+def pow2(exponent):
+    return 1 << exponent
+
+
+def get_freq(hash_value):
+    packed = struct.pack(">q", hash_value)
+    return struct.unpack(">d", packed)[0]
+
+
+class HiPS3DArray:
+
+    def __init__(self, directory_or_url, level=None):
+
+        self._cache = {}
+
+        self._directory_or_url = directory_or_url
+
+        self._is_url = is_url(directory_or_url)
+
+        self._properties = load_properties(directory_or_url)
+
+        assert self._properties["dataproduct_type"] == "spectral-cube"
+
+        self._tile_width = int(self._properties["hips_tile_width"])
+        self._tile_depth = int(self._properties["hips_tile_depth"])
+
+        self._order = int(self._properties["hips_order"])
+        self._order_freq = int(self._properties["hips_order_freq"])
+
+        # FIXME: for now assume minimum order is 0
+
+        self._level = self._order if level is None else level
+        self._level_freq = self._level + (self._order_freq - self._order)
+
+        self._tile_format = self._properties["hips_tile_format"]
+
+        self._nside = level_to_nside(self._level)
+        self._hp = HEALPix(nside=self._nside, frame="icrs", order="nested")
+        self._cdelt = 45 / self._tile_width / 2**self._level * 2**0.5
+
+        image_size = 5 * self._nside * self._tile_width
+
+        # For the image depth we could in principe do whole spectral domain but
+        # this would make too many chunks for dask so we have to be more
+        # sensible
+
+        # NOTE: em_min is given as wav but is minimum frequency
+
+        self._freq_min = (float(self._properties["em_min"]) * u.m).to_value(u.Hz, u.spectral())
+        self._freq_max = (float(self._properties["em_max"]) * u.m).to_value(u.Hz, u.spectral())
+
+        # Now determine what the indices would be for this at the given spectral order
+
+        self._index_min = freq2pix(self._level_freq, self._freq_min)
+        self._index_max = freq2pix(self._level_freq, self._freq_max) + 1
+
+        image_depth = (self._index_max - self._index_min) * self._tile_depth
+
+        # FIXME: make WCS 3D
+
+        self.wcs = WCS(naxis=2)
+        self.wcs.wcs.ctype = "RA---HPX", "DEC--HPX"
+        self.wcs.wcs.crval = 0, 0
+        self.wcs.wcs.cdelt = self._cdelt, self._cdelt
+        self.wcs.wcs.crpix = image_size / 2, image_size / 2
+        self.wcs.wcs.crota = 0, 45
+        self.wcs.wcs.set()
+
+        self.shape = (image_depth, image_size, image_size)
+        self.dtype = float
+        self.ndim = 2
+
+        self.chunksize = (self._tile_depth, self._tile_width, self._tile_width)
+
+        self._nan = np.nan * np.ones(self.chunksize, dtype=self.dtype)
+
+    def __getitem__(self, item):
+
+        # For now assume item is a list of slices. Find
+
+        ispec = (item[0].start + item[0].stop) // 2
+        imid = (item[1].start + item[1].stop) // 2
+        jmid = (item[2].start + item[2].stop) // 2
+
+        # Convert pixel coordinates to HEALPix indices
+
+        spatial_index = self._hp.skycoord_to_healpix(self.wcs.pixel_to_world(imid, jmid))
+
+        if spatial_index == -1:
+            return self._nan
+
+        # Get spectral index
+
+        spectral_index = ispec // self._tile_depth + self._index_min
+
+        return self._get_tile(
+            level=self._level, spatial_index=spatial_index, spectral_index=spectral_index
+        )
+
+    @functools.lru_cache(maxsize=128)
+    def _get_tile(self, *, level, spatial_index, spectral_index):
+
+        filename_or_url = tile_filename_3d(
+            spatial_level=self._level,
+            spectral_level=self._level_freq,
+            spatial_index=spatial_index,
+            spectral_index=spectral_index,
+            output_directory=self._directory_or_url,
+            extension="fits",
+        )
+
+        if self._is_url:
+            try:
+                filename = download_file(filename_or_url, cache=True)
+            except urllib.error.HTTPError:
+                return self._nan
+        elif not os.path.exists(filename_or_url):
+            return self._nan
+        else:
+            filename = filename_or_url
+
+        with fits.open(filename) as hdulist:
+
+            hdu = hdulist[0]
+            data = hdu.data
+
+            if data.shape != self.chunksize:
+
+                # Need to add padding
+
+                before = (hdu.header["TRIM3"], hdu.header["TRIM2"], hdu.header["TRIM1"])
+                after = [
+                    (c - s - b)
+                    for (c, s, b) in zip(self.chunksize, data.shape, before, strict=False)
+                ]
+
+                data = np.pad(data, list(zip(before, after, strict=False)), constant_values=np.nan)
+
+            data = data[:, ::-1, :]
+
+            return data
+
+
+def hips_as_dask(directory_or_url, level=None):
+    array_wrapper = HiPSArray(directory_or_url, level=level)
+    return da.from_array(
+        array_wrapper,
+        chunks=array_wrapper.chunksize,
+        name=str(uuid.uuid4()),
+    )
+
+
+def hips3d_as_dask(directory_or_url, level=None):
+    array_wrapper = HiPS3DArray(directory_or_url, level=level)
+    return da.from_array(
+        array_wrapper,
+        chunks=array_wrapper.chunksize,
+        name=str(uuid.uuid4()),
+    )

reproject/hips/high_level.py

@@ -22,7 +22,9 @@
 from ..utils import as_transparent_rgb, is_jpeg, is_png, parse_input_data
 from ..wcs_utils import has_celestial, pixel_scale
 from .utils import (
+    load_properties,
     make_tile_folders,
+    save_properties,
     tile_filename,
     tile_header,
 )
@@ -403,21 +405,6 @@ def save_index(directory):
         f.write(INDEX_HTML)
 
 
-def save_properties(directory, properties):
-    with open(os.path.join(directory, "properties"), "w") as f:
-        for key, value in properties.items():
-            f.write(f"{key:20s} = {value}\n")
-
-
-def load_properties(directory):
-    properties = {}
-    with open(os.path.join(directory, "properties")) as f:
-        for line in f:
-            key, value = line.split("=")
-            properties[key.strip()] = value.strip()
-    return properties
-
-
 def coadd_hips(input_directories, output_directory):
     """
     Given multiple HiPS directories, combine these into a single HiPS.