Merge pull request #6 from pytroll/master

Bring my fork up to date
adybbroe · Apr 11, 2019 · 8f6ffba · 8f6ffba
2 parents 3e4f423 + 35fbc9f
commit 8f6ffba
Show file tree

Hide file tree

Showing 16 changed files with 328 additions and 58 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,3 +1,55 @@
+## Version 0.14.0 (2019/04/09)
+
+### Issues Closed
+
+* [Issue 698](https://github.com/pytroll/satpy/issues/698) - Read WKT geotiff
+* [Issue 692](https://github.com/pytroll/satpy/issues/692) - sdr_viirs_l1b reader fails in 0.13, recent master, Works with version 0.12.0 ([PR 693](https://github.com/pytroll/satpy/pull/693))
+* [Issue 683](https://github.com/pytroll/satpy/issues/683) - Question: Change image size when saving with satpy.save_dataset ([PR 691](https://github.com/pytroll/satpy/pull/691))
+* [Issue 681](https://github.com/pytroll/satpy/issues/681) - incorrect data offset in HSD files ([PR 689](https://github.com/pytroll/satpy/pull/689))
+* [Issue 666](https://github.com/pytroll/satpy/issues/666) - Add drawing of lat lon graticules when saving dataset ([PR 668](https://github.com/pytroll/satpy/pull/668))
+* [Issue 646](https://github.com/pytroll/satpy/issues/646) - Add 'demo' subpackage for accessing example data ([PR 686](https://github.com/pytroll/satpy/pull/686))
+* [Issue 528](https://github.com/pytroll/satpy/issues/528) - Support dask version of PySpectral ([PR 529](https://github.com/pytroll/satpy/pull/529))
+* [Issue 511](https://github.com/pytroll/satpy/issues/511) - Add/update documentation about composites and compositors ([PR 705](https://github.com/pytroll/satpy/pull/705))
+
+In this release 8 issues were closed.
+
+### Pull Requests Merged
+
+#### Bugs fixed
+
+* [PR 700](https://github.com/pytroll/satpy/pull/700) - Mask out invalid values in the precipitation probability product
+* [PR 693](https://github.com/pytroll/satpy/pull/693) - Fix VIIRS SDR reading of visible channels at nighttime ([692](https://github.com/pytroll/satpy/issues/692))
+* [PR 689](https://github.com/pytroll/satpy/pull/689) - Fix Himawari HSD reader's incorrect header information ([681](https://github.com/pytroll/satpy/issues/681))
+* [PR 688](https://github.com/pytroll/satpy/pull/688) - Fix offset correction in seviri_l1b_hrit
+* [PR 685](https://github.com/pytroll/satpy/pull/685) - Fix bug in Scene.resample causing AssertionError
+* [PR 677](https://github.com/pytroll/satpy/pull/677) - Fix MultiScene save_animation when distributed isn't installed
+* [PR 675](https://github.com/pytroll/satpy/pull/675) - Do not pass `filter_parameters` to the filehandler creation
+
+#### Features added
+
+* [PR 691](https://github.com/pytroll/satpy/pull/691) - Add Scene.aggregate method (python 3 only) ([683](https://github.com/pytroll/satpy/issues/683))
+* [PR 686](https://github.com/pytroll/satpy/pull/686) - Add demo subpackage to simplify test data download ([646](https://github.com/pytroll/satpy/issues/646))
+* [PR 676](https://github.com/pytroll/satpy/pull/676) - Feature add nightfog modis
+* [PR 674](https://github.com/pytroll/satpy/pull/674) - Use platform ID to choose the right reader for AVHRR GAC data
+* [PR 671](https://github.com/pytroll/satpy/pull/671) - Add satellite position to dataset attributes (seviri_l1b_hrit)
+* [PR 669](https://github.com/pytroll/satpy/pull/669) - Add ocean-color for viirs and modis
+* [PR 668](https://github.com/pytroll/satpy/pull/668) - Add grid/graticules to add_overlay function. ([666](https://github.com/pytroll/satpy/issues/666))
+* [PR 665](https://github.com/pytroll/satpy/pull/665) - Add reader for VIIRS Active Fires
+* [PR 645](https://github.com/pytroll/satpy/pull/645) - Reader for the SAR OCN L2 wind product in SAFE format.
+* [PR 565](https://github.com/pytroll/satpy/pull/565) - Add reader for FY-3 VIRR (virr_l1b)
+* [PR 529](https://github.com/pytroll/satpy/pull/529) - Add dask support to NIRReflectance modifier ([528](https://github.com/pytroll/satpy/issues/528))
+
+#### Documentation changes
+
+* [PR 707](https://github.com/pytroll/satpy/pull/707) - Add ABI Meso demo data case and clean up documentation
+* [PR 705](https://github.com/pytroll/satpy/pull/705) - Document composites ([511](https://github.com/pytroll/satpy/issues/511))
+* [PR 701](https://github.com/pytroll/satpy/pull/701) - Clarify release instructions
+* [PR 699](https://github.com/pytroll/satpy/pull/699) - Rename SatPy to Satpy throughout documentation
+* [PR 673](https://github.com/pytroll/satpy/pull/673) - Add information about GDAL_CACHEMAX to FAQ
+
+In this release 23 pull requests were closed.
+
+
 ## Version 0.13.0 (2019/03/18)
 
 ### Issues Closed

diff --git a/doc/source/api/satpy.demo.rst b/doc/source/api/satpy.demo.rst
@@ -1,18 +1,6 @@
 satpy.demo package
 ==================
 
-Submodules
-----------
-
-satpy.demo.google\_cloud\_platform module
------------------------------------------
-
-.. automodule:: satpy.demo.google_cloud_platform
-    :members:
-    :undoc-members:
-    :show-inheritance:
-
-
 Module contents
 ---------------
 

diff --git a/doc/source/api/satpy.readers.rst b/doc/source/api/satpy.readers.rst
@@ -365,7 +365,7 @@ satpy.readers.viirs\_compact module
     :show-inheritance:
 
 satpy.readers.viirs\_edr\_active\_fires module
---------------------------------------
+----------------------------------------------
 
 .. automodule:: satpy.readers.viirs_edr_active_fires
     :members:
@@ -396,6 +396,14 @@ satpy.readers.viirs\_sdr module
     :undoc-members:
     :show-inheritance:
 
+satpy.readers.virr\_l1b module
+------------------------------
+
+.. automodule:: satpy.readers.virr_l1b
+    :members:
+    :undoc-members:
+    :show-inheritance:
+
 satpy.readers.xmlformat module
 ------------------------------
 

diff --git a/doc/source/conf.py b/doc/source/conf.py
@@ -60,7 +60,8 @@ def __getattr__(cls, name):
 # Add any Sphinx extension module names here, as strings. They can be extensions
 # coming with Sphinx (named 'sphinx.ext.*') or your custom ones.
 extensions = ['sphinx.ext.autodoc', 'sphinx.ext.intersphinx', 'sphinx.ext.todo', 'sphinx.ext.coverage',
-              'sphinx.ext.doctest', 'sphinx.ext.napoleon', 'sphinx.ext.autosummary', 'doi_role']
+              'sphinx.ext.doctest', 'sphinx.ext.napoleon', 'sphinx.ext.autosummary', 'doi_role',
+              'sphinx.ext.viewcode']
 
 # Add any paths that contain templates here, relative to this directory.
 templates_path = ['_templates']

diff --git a/doc/source/data_download.rst b/doc/source/data_download.rst
@@ -2,8 +2,11 @@ Downloading Data
 ================
 
 One of the main features of Satpy is its ability to read various satellite
-data formats. However, it does not currently provide any functionality for
-downloading data from any remote sources. Satpy assumes all data is available
+data formats. However, it currently only provides limited methods for
+downloading data from remote sources and these methods are limited to demo
+data for `Pytroll examples <https://github.com/pytroll/pytroll-examples>`_.
+See the examples and the :mod:`~satpy.demo` API documentation for details.
+Otherwise, Satpy assumes all data is available
 through the local system, either as a local directory or network
 mounted file systems. Certain readers that use ``xarray`` to open data files
 may be able to load files from remote systems by using OpenDAP or similar

diff --git a/satpy/composites/__init__.py b/satpy/composites/__init__.py
@@ -49,6 +49,16 @@
 from satpy.writers import get_enhanced_image
 from satpy import CHUNK_SIZE
 
+try:
+    from pyspectral.near_infrared_reflectance import Calculator
+except ImportError:
+    Calculator = None
+try:
+    from pyorbital.astronomy import sun_zenith_angle
+except ImportError:
+    sun_zenith_angle = None
+
+
 LOG = logging.getLogger(__name__)
 
 
@@ -550,7 +560,7 @@ def __call__(self, projectables, optional_datasets=None, **info):
         self._init_refl3x(projectables)
         _nir, _ = projectables
         refl = self._get_reflectance(projectables, optional_datasets) * 100
-        proj = xr.DataArray(refl.filled(np.nan), dims=_nir.dims,
+        proj = xr.DataArray(refl, dims=_nir.dims,
                             coords=_nir.coords, attrs=_nir.attrs)
 
         proj.attrs['units'] = '%'
@@ -560,12 +570,9 @@ def __call__(self, projectables, optional_datasets=None, **info):
 
     def _init_refl3x(self, projectables):
         """Initiate the 3.x reflectance derivations."""
-        try:
-            from pyspectral.near_infrared_reflectance import Calculator
-        except ImportError:
+        if not Calculator:
             LOG.info("Couldn't load pyspectral")
-            raise
-
+            raise ImportError("No module named pyspectral.near_infrared_reflectance")
         _nir, _tb11 = projectables
         self._refl3x = Calculator(_nir.attrs['platform_name'], _nir.attrs['sensor'], _nir.attrs['name'])
 
@@ -588,9 +595,10 @@ def _get_reflectance(self, projectables, optional_datasets):
 
         # Check if the sun-zenith angle was provided:
         if sun_zenith is None:
-            from pyorbital.astronomy import sun_zenith_angle as sza
+            if sun_zenith_angle is None:
+                raise ImportError("No module named pyorbital.astronomy")
             lons, lats = _nir.attrs["area"].get_lonlats_dask(CHUNK_SIZE)
-            sun_zenith = sza(_nir.attrs['start_time'], lons, lats)
+            sun_zenith = sun_zenith_angle(_nir.attrs['start_time'], lons, lats)
 
         return self._refl3x.reflectance_from_tbs(sun_zenith, _nir, _tb11, tb_ir_co2=tb13_4)
 
@@ -610,7 +618,6 @@ def __call__(self, projectables, optional_datasets=None, **info):
         # needs to be derived first in order to get the emissive part.
         _ = self._get_reflectance(projectables, optional_datasets)
         _nir, _ = projectables
-        raise NotImplementedError("This compositor wasn't fully converted to dask yet.")
         proj = xr.DataArray(self._refl3x.emissive_part_3x(), attrs=_nir.attrs,
                             dims=_nir.dims, coords=_nir.coords)
 

diff --git a/satpy/demo/__init__.py b/satpy/demo/__init__.py
@@ -22,17 +22,22 @@
 for how the data is downloaded (via the ``method`` keyword argument)
 including:
 
-- gcsfs: Download data from a public google cloud storage bucket using the
+- gcsfs:
+    Download data from a public google cloud storage bucket using the
     ``gcsfs`` package.
-- unidata_thredds: Access data using OpenDAP or similar method from Unidata's
+- unidata_thredds:
+    Access data using OpenDAP or similar method from Unidata's
     public THREDDS server
     (https://thredds.unidata.ucar.edu/thredds/catalog.html).
-- uwaos_thredds: Access data using OpenDAP or similar method from the
+- uwaos_thredds:
+    Access data using OpenDAP or similar method from the
     University of Wisconsin - Madison's AOS department's THREDDS server.
-- http: A last resort download method when nothing else is available of a
+- http:
+    A last resort download method when nothing else is available of a
     tarball or zip file from one or more servers available to the Satpy
     project.
-- uw_arcdata: A network mount available on many servers at the Space Science
+- uw_arcdata:
+    A network mount available on many servers at the Space Science
     and Engineering Center (SSEC) at the University of Wisconsin - Madison.
     This is method is mainly meant when tutorials are taught at the SSEC
     using a Jupyter Hub server.
@@ -51,7 +56,7 @@
 LOG = logging.getLogger(__name__)
 
 
-def makedirs(directory, exist_ok=False):
+def _makedirs(directory, exist_ok=False):
     """Python 2.7 friendly os.makedirs.
 
     After python 2.7 is dropped, just use `os.makedirs` with `existsok=True`.
@@ -64,7 +69,7 @@ def makedirs(directory, exist_ok=False):
 
 
 def get_us_midlatitude_cyclone_abi(base_dir='.', method=None, force=False):
-    """Get GOES-16 ABI (CONUS sector) data from March 14th 00:00Z.
+    """Get GOES-16 ABI (CONUS sector) data from 2019-03-14 00:00Z.
 
     Args:
         base_dir (str): Base directory for downloaded files.
@@ -75,17 +80,75 @@ def get_us_midlatitude_cyclone_abi(base_dir='.', method=None, force=False):
             the local system. Warning: May delete non-demo files stored in
             download directory.
 
+    Total size: ~110MB
+
     """
     if method is None:
         method = 'gcsfs'
     if method not in ['gcsfs']:
         raise NotImplementedError("Demo data download method '{}' not "
                                   "implemented yet.".format(method))
 
-    from .google_cloud_platform import get_bucket_files
+    from ._google_cloud_platform import get_bucket_files
     patterns = ['gs://gcp-public-data-goes-16/ABI-L1b-RadC/2019/073/00/*0002*.nc']
     subdir = os.path.join(base_dir, 'abi_l1b', '20190314_us_midlatitude_cyclone')
-    makedirs(subdir, exist_ok=True)
+    _makedirs(subdir, exist_ok=True)
     filenames = get_bucket_files(patterns, subdir, force=force)
-    assert len(filenames) == 16, "Not all ABI files could be downloaded"
+    assert len(filenames) == 16, "Not all files could be downloaded"
+    return filenames
+
+
+def get_hurricane_florence_abi(base_dir='.', method=None, force=False,
+                               channels=range(1, 17), num_frames=10):
+    """Get GOES-16 ABI (Meso sector) data from 2018-09-11 13:00Z to 17:00Z.
+
+    Args:
+        base_dir (str): Base directory for downloaded files.
+        method (str): Force download method for the data if not already cached.
+            Allowed options are: 'gcsfs'. Default of ``None`` will
+            choose the best method based on environment settings.
+        force (bool): Force re-download of data regardless of its existence on
+            the local system. Warning: May delete non-demo files stored in
+            download directory.
+        channels (list): Channels to include in download. Defaults to all
+            16 channels.
+        num_frames (int or slice): Number of frames to download. Maximum
+            240 frames. Default 10 frames.
+
+    Size per frame (all channels): ~15MB
+
+    Total size (default 10 frames, all channels): ~124MB
+
+    Total size (240 frames, all channels): ~3.5GB
+
+    """
+    if method is None:
+        method = 'gcsfs'
+    if method not in ['gcsfs']:
+        raise NotImplementedError("Demo data download method '{}' not "
+                                  "implemented yet.".format(method))
+    if isinstance(num_frames, (int, float)):
+        frame_slice = slice(0, num_frames)
+    else:
+        frame_slice = num_frames
+
+    from ._google_cloud_platform import get_bucket_files
+
+    patterns = []
+    for channel in channels:
+        # patterns += ['gs://gcp-public-data-goes-16/ABI-L1b-RadM/2018/254/1[3456]/'
+        #              '*C{:02d}*s20182541[3456]*.nc'.format(channel)]
+        patterns += [(
+            'gs://gcp-public-data-goes-16/ABI-L1b-RadM/2018/254/13/*RadM1*C{:02d}*s201825413*.nc'.format(channel),
+            'gs://gcp-public-data-goes-16/ABI-L1b-RadM/2018/254/14/*RadM1*C{:02d}*s201825414*.nc'.format(channel),
+            'gs://gcp-public-data-goes-16/ABI-L1b-RadM/2018/254/15/*RadM1*C{:02d}*s201825415*.nc'.format(channel),
+            'gs://gcp-public-data-goes-16/ABI-L1b-RadM/2018/254/16/*RadM1*C{:02d}*s201825416*.nc'.format(channel),
+        )]
+    subdir = os.path.join(base_dir, 'abi_l1b', '20180911_hurricane_florence_abi_l1b')
+    _makedirs(subdir, exist_ok=True)
+    filenames = get_bucket_files(patterns, subdir, force=force, pattern_slice=frame_slice)
+
+    actual_slice = frame_slice.indices(240)  # 240 max frames
+    num_frames = int((actual_slice[1] - actual_slice[0]) / actual_slice[2])
+    assert len(filenames) == len(channels) * num_frames, "Not all files could be downloaded"
     return filenames
diff --git a/satpy/demo/google_cloud_platform.py → satpy/demo/_google_cloud_platform.py b/satpy/demo/google_cloud_platform.py → satpy/demo/_google_cloud_platform.py
@@ -41,18 +41,24 @@ def is_google_cloud_instance():
         return False
 
 
-def get_bucket_files(glob_pattern, base_dir, force=False):
+def get_bucket_files(glob_pattern, base_dir, force=False, pattern_slice=slice(None)):
     """Helper function to download files from Google Cloud Storage.
 
     Args:
         glob_pattern (str or list): Glob pattern string or series of patterns
             used to search for on Google Cloud Storage. The pattern should
-            include the "gs://" protocol prefix.
+            include the "gs://" protocol prefix. If a list of lists, then the
+            results of each sublist pattern are concatenated and the result is
+            treated as one pattern result. This is important for things like
+            ``pattern_slice`` and complicated glob patterns not supported by
+            GCP.
         base_dir (str): Root directory to place downloaded files on the local
             system.
         force (bool): Force re-download of data regardless of its existence on
             the local system. Warning: May delete non-demo files stored in
             download directory.
+        pattern_slice (slice): Slice object to limit the number of files
+            returned by each glob pattern.
 
     """
     if gcsfs is None:
@@ -67,16 +73,25 @@ def get_bucket_files(glob_pattern, base_dir, force=False):
     fs = gcsfs.GCSFileSystem(token='anon')
     filenames = []
     for gp in glob_pattern:
-        for fn in fs.glob(gp):
+        # handle multiple glob patterns being treated as one pattern
+        # for complicated patterns that GCP can't handle
+        if isinstance(gp, str):
+            glob_results = list(fs.glob(gp))
+        else:
+            # flat list of results
+            glob_results = [fn for pat in gp for fn in fs.glob(pat)]
+
+        for fn in glob_results[pattern_slice]:
             ondisk_fn = os.path.basename(fn)
             ondisk_pathname = os.path.join(base_dir, ondisk_fn)
             filenames.append(ondisk_pathname)
-            LOG.info("Downloading: {}".format(ondisk_pathname))
 
             if force and os.path.isfile(ondisk_pathname):
                 os.remove(ondisk_pathname)
             elif os.path.isfile(ondisk_pathname):
+                LOG.info("Found existing: {}".format(ondisk_pathname))
                 continue
+            LOG.info("Downloading: {}".format(ondisk_pathname))
             fs.get('gs://' + fn, ondisk_pathname)
 
     if not filenames:

diff --git a/satpy/etc/areas.yaml b/satpy/etc/areas.yaml
@@ -399,8 +399,8 @@ seviri_0deg:
     height: 3712
     width: 3712
   area_extent:
-    lower_left_xy: [-5570248.477339261, -5567248.074173444]
-    upper_right_xy: [5567248.074173444, 5570248.477339261]
+    lower_left_xy: [-5570248.686685662, -5567248.28340708]
+    upper_right_xy: [5567248.28340708, 5570248.686685662]
 seviri_iodc:
   description: Full globe MSG image 41.5 degrees
   projection:
@@ -413,8 +413,8 @@ seviri_iodc:
     height: 3712
     width: 3712
   area_extent:
-    lower_left_xy: [-5570248.477339261, -5567248.074173444]
-    upper_right_xy: [5567248.074173444, 5570248.477339261]
+    lower_left_xy: [-5570248.686685662, -5567248.28340708]
+    upper_right_xy: [5567248.28340708, 5570248.686685662]
 msg_resample_area:
   description: Full globe MSG image 20.75 degrees
   projection:

diff --git a/satpy/readers/seviri_l1b_hrit.py b/satpy/readers/seviri_l1b_hrit.py
@@ -213,9 +213,9 @@ def _find_navigation_coefs(self):
         # Find index of interval enclosing the nominal timestamp of the scan
         time = np.datetime64(self.prologue['ImageAcquisition']['PlannedAcquisitionTime']['TrueRepeatCycleStart'])
         intervals_tstart = self.prologue['SatelliteStatus']['Orbit']['OrbitPolynomial']['StartTime'][0].astype(
-            'datetime64')
+            'datetime64[us]')
         intervals_tend = self.prologue['SatelliteStatus']['Orbit']['OrbitPolynomial']['EndTime'][0].astype(
-            'datetime64')
+            'datetime64[us]')
         try:
             return np.where(np.logical_and(time >= intervals_tstart, time < intervals_tend))[0][0]
         except IndexError: