panoptes-data 0.2.3__tar.gz → 0.3.0__tar.gz

{panoptes_data-0.2.3 → panoptes_data-0.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.4
+Metadata-Version: 2.5
 Name: panoptes-data
-Version: 0.2.3
+Version: 0.3.0
 Summary: Tools for working with PANOPTES data.
 Project-URL: Documentation, https://projectpanoptes.org/
 Project-URL: Source, https://github.com/panoptes/panoptes-data/

{panoptes_data-0.2.3 → panoptes_data-0.3.0}/src/panoptes/data/observations.py

@@ -13,6 +13,10 @@ from panoptes.utils.images import fits as fits_utils
 
 warnings.filterwarnings('ignore', category=FITSFixedWarning)
 
+# The metadata columns this class actually needs. Everything else, including
+# every URL column, is optional -- see `ObservationInfo.public_urls`.
+REQUIRED_COLUMNS = ('time', 'uid')
+
 
 class ObservationInfo:
     """A container class for information about an Observation."""
@@ -49,7 +53,23 @@ class ObservationInfo:
             self.meta = dict()
 
         self.image_metadata = self.get_metadata(query=image_query)
-        self.image_list = self.image_metadata.public_url.values
+        self.image_list = self.get_image_list()
+
+    @property
+    def public_urls(self):
+        """The browsable URLs carried by the metadata, if any.
+
+        Which ``*_url`` columns exist varies between records: ``public_url``
+        and ``raw_url`` on some, ``uploaded_public_url``, ``fits_public_url``
+        and ``jpg_public_url`` on others. They are for display and lookup only
+        -- the raw frame locations come from `get_image_list` -- so a record
+        with no URL column at all still works.
+
+        Returns:
+            A DataFrame of whichever ``*_url`` columns are present, which may
+            have no columns.
+        """
+        return self.image_metadata.filter(regex=r'_url$')
 
     def get_image_cutout(self, data=None, coords=None, box_size=None, *args, **kwargs):
         """Gets a Cutout2D object for the given coords and box_size."""
@@ -72,6 +92,13 @@ class ObservationInfo:
         metadata_url = f'{self._settings.img_metadata_url.unicode_string()}?sequence_id={self.sequence_id}'
         images_df = pd.read_csv(metadata_url)
 
+        missing = [col for col in REQUIRED_COLUMNS if col not in images_df.columns]
+        if missing:
+            raise ValueError(
+                f'Metadata for {self.sequence_id} is missing required '
+                f'column(s) {missing}; got {sorted(images_df.columns)}'
+            )
+
         # Set a time index.
         images_df.time = pd.to_datetime(images_df.time)
         images_df = images_df.set_index(['time']).sort_index()
@@ -82,13 +109,18 @@ class ObservationInfo:
         return images_df
 
     def get_image_list(self, bucket: str | None = None, file_ext: str = '.fits.fz'):
-        """Get the images for the observation.
+        """Get the URLs of the raw images for the observation.
+
+        The URLs are built from each image's ``uid`` rather than read from a URL
+        column, which is not dependable (`public_urls`). The ``uid`` is the
+        archive path with underscores for separators, so the raw frame's
+        location follows from it and the bucket.
 
         Args:
              bucket: The bucket where the images are stored.
              file_ext: The file extension of the images to retrieve.
         Returns:
-            A pandas DataFrame containing the images for the observation.
+            A list of URLs, one per image in the metadata.
         """
         url_base = self._settings.img_base_url.unicode_string()
         bucket = bucket or self._settings.img_bucket

{panoptes_data-0.2.3 → panoptes_data-0.3.0}/src/panoptes/data/utils/cli/main.py

@@ -42,6 +42,7 @@ def download(sequence_id: str | None = typer.Argument(..., help='Sequence ID for
 
     except Exception as e:
         print(f'[red]Error downloading images for {sequence_id}: {e}')
+        raise typer.Exit(code=1) from e
 
     return local_files
 
@@ -84,6 +85,7 @@ def get_metadata(
             print(f'[green]Metadata saved to {output_fn}')
         except Exception as e:
             print(f'[red]Error downloading metadata for {sequence_id}: {e}')
+            raise typer.Exit(code=1) from e
     else:
         if unit_id is None:
             print('[red]Must provide a unit_id if not providing a sequence_id.')
@@ -120,7 +122,8 @@ def get_metadata(
             pd.concat(dfs).to_csv(output_fn)
 
         except ValueError as e:
-            print(f'Error downloading metadata for {unit_id}: {e}', fg='red')
+            print(f'[red]Error downloading metadata for {unit_id}: {e}')
+            raise typer.Exit(code=1) from e
 
     print(f'Metadata saved to [green]{output_fn}')