v0.1.5

## [0.1.5] - 2024-06-05

### Fixed

- Better management of cloud mask and error messages.
- `groupby_date` is now performed after processing the cloudmask in order to ensure a better compatibility.

Author: nkarasiak

.github/workflows/pytest-dev.yaml

@@ -11,7 +11,6 @@ jobs:
         os:
           - ubuntu-latest
         python-version:
-          - "3.10"
           - "3.11"
           - "3.12"
     runs-on:  ${{ matrix.os }}

.github/workflows/pytest-prod.yaml

@@ -11,8 +11,8 @@ jobs:
         os:
           - ubuntu-latest
           - windows-latest
+          - macos-latest
         python-version:
-          - "3.10"
           - "3.11"
           - "3.12"
     runs-on:  ${{ matrix.os }}

CHANGELOG.md

@@ -4,6 +4,13 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.1.5] - 2024-06-05
+
+### Fixed
+
+- Better management of cloud mask and error messages.
+- `groupby_date` is now performed after processing the cloudmask in order to ensure a better compatibility.
+
 ## [0.1.4] - 2024-05-23
 
 ### Fixed

earthdaily/__init__.py

@@ -5,4 +5,4 @@
 # to hide warnings from rioxarray or nano seconds conversion
 warnings.filterwarnings("ignore")
 
-__version__ = "0.1.4"
+__version__ = "0.1.5"

earthdaily/accessor/__init__.py

@@ -117,6 +117,29 @@ def whittaker(
             max_iter=max_iter,
         )["index"]
 
+    def sel_nearest_dates(
+        self,
+        target: (xr.Dataset, xr.DataArray),
+        max_delta: int = 0,
+        method: str = "nearest",
+        return_target: bool = False,
+    ):
+        src_time = self._obj.sel(time=target.time.dt.date, method=method).time.dt.date
+        target_time = target.time.dt.date
+        pos = np.abs(src_time.data - target_time.data)
+        pos = [
+            src_time.isel(time=i).time.values
+            for i, j in enumerate(pos)
+            if j.days <= max_delta
+        ]
+        pos = np.unique(pos)
+        if return_target:
+            method_convert = {"bfill": "ffill", "ffill": "bfill", "nearest": "nearest"}
+            return self._obj.sel(time=pos), target.sel(
+                time=pos, method=method_convert[method]
+            )
+        return self._obj.sel(time=pos)
+
 
 @xr.register_dataset_accessor("ed")
 class EarthDailyAccessorDataset:
@@ -199,6 +222,9 @@ def _auto_mapper(self):
             "rededge1": "RE1",
             "rededge2": "RE2",
             "rededge3": "RE3",
+            "rededge70": "RE1",
+            "rededge74": "RE2",
+            "rededge78": "RE3",
             "nir": "N",
             "nir08": "N2",
             "watervapor": "WV",

earthdaily/earthdatastore/__init__.py

@@ -641,9 +641,8 @@ def datacube(
 
         if mask_with:
             if mask_with not in mask._available_masks:
-                logging.log(
-                    level=logging.INFO,
-                    msg=f"Specified mask '{mask_with}' is not available. Currently available masks provider are : {mask._available_masks}",
+                raise NotImplementedError(
+                    f"Specified mask '{mask_with}' is not available. Available masks providers are : {mask._available_masks}"
                 )
 
             elif mask_with == "ag_cloud_mask":
@@ -686,6 +685,10 @@ def datacube(
                     "No cross calibration coefficient available for the specified collections."
                 )
 
+        groupby_date_sensor_cube = groupby_date
+        if mask_with and groupby_date:
+            groupby_date_sensor_cube = None
+
         xr_datacube = datacube(
             items,
             intersects=intersects,
@@ -694,11 +697,10 @@ def datacube(
             common_band_names=common_band_names,
             cross_calibration_items=xcal_items,
             properties=properties,
-            groupby_date=groupby_date,
+            groupby_date=groupby_date_sensor_cube,
             **kwargs,
         )
         if mask_with:
-            groupby_date_mask = "max" if isinstance(groupby_date, str) else None
             if clear_cover and mask_statistics is False:
                 mask_statistics = True
             mask_kwargs = dict(mask_statistics=mask_statistics)
@@ -708,7 +710,7 @@ def datacube(
                     acm_items,
                     intersects=intersects,
                     bbox=bbox,
-                    groupby_date=groupby_date_mask,
+                    groupby_date=None,
                     geobox=xr_datacube.odc.geobox
                     if hasattr(xr_datacube, "odc")
                     else None,
@@ -728,7 +730,7 @@ def datacube(
 
                 clouds_datacube = datacube(
                     items,
-                    groupby_date=groupby_date_mask,
+                    groupby_date=None,
                     intersects=intersects,
                     bbox=bbox,
                     assets=[mask_assets],
@@ -750,6 +752,29 @@ def datacube(
 
             if clear_cover:
                 xr_datacube = mask.filter_clear_cover(xr_datacube, clear_cover)
+        if groupby_date and mask_with:
+            grouped_coords = []
+            # for coords using only time dimensions like clear_pixels, keeping the max
+            for coord in xr_datacube.coords:
+                if coord in ("x", "y", "time"):
+                    continue
+                if (
+                    len(xr_datacube[coord].dims) == 1
+                    and xr_datacube[coord].dims[0] == "time"
+                ):
+                    grouped_coords.append(
+                        xr_datacube[coord]
+                        .groupby("time.date", squeeze=True)
+                        .max()
+                        .rename(dict(date="time"))
+                    )
+
+            xr_datacube = xr_datacube.groupby("time.date", restore_coord_dims=True)
+            xr_datacube = getattr(xr_datacube, groupby_date)().rename(dict(date="time"))
+            for grouped_coord in grouped_coords:
+                xr_datacube = xr_datacube.assign_coords(
+                    {grouped_coord.name: grouped_coord}
+                )
         return xr_datacube
 
     def _update_search_for_assets(self, assets):

earthdaily/earthdatastore/cube_utils/__init__.py

@@ -23,13 +23,14 @@ def _impl(self, *args, **kwargs):
         mask_with = kwargs.get("mask_with", None)
         if isinstance(mask_with, list):
             kwargs.pop("mask_with")
-            for mask in mask_with:
+            for idx, mask in enumerate(mask_with):
                 try:
                     datacube = method(self, mask_with=mask, *args, **kwargs)
                     break
-                except Warning:
+                except Warning as E:
                     # if warning about no items for ag_cloud_mask for example
-                    continue
+                    if idx + 1 == len(mask_with):
+                        raise E
         else:
             datacube = method(self, *args, **kwargs)
         return datacube

earthdaily/earthdatastore/cube_utils/geometry_manager.py

@@ -51,13 +51,21 @@ def _unknow_geometry_to_geodataframe(self, geometry):
                 return gpd.read_file(geometry, driver="GeoJson", crs="EPSG:4326")
             except:
                 try:
-                    return gpd.GeoDataFrame.from_features(geometry, crs="EPSG:4326")
+                    return gpd.read_file(
+                        json.dumps(geometry), driver="GeoJson", crs="EPSG:4326"
+                    )
                 except:
-                    if "type" in geometry:
-                        geom = shapely.__dict__[geometry["type"]](
-                            [geometry["coordinates"][0]]
-                        )
-                        return gpd.GeoDataFrame(geometry=[geom], crs="EPSG:4326")
+                    pass
+
+            try:
+                return gpd.GeoDataFrame.from_features(geometry, crs="EPSG:4326")
+            except:
+                if "type" in geometry:
+                    geom = shapely.__dict__[geometry["type"]](
+                        [geometry["coordinates"][0]]
+                    )
+                    return gpd.GeoDataFrame(geometry=[geom], crs="EPSG:4326")
+
         elif isinstance(geometry, gpd.GeoSeries):
             self.input_type = "GeoSeries"
 

earthdaily/earthdatastore/mask/__init__.py

@@ -162,7 +162,6 @@ def compute_clear_coverage(
                 )
             }
         )
-        # self._obj = self._obj.reset_coords(names=['clear_pixels','clear_percent'])
 
         return self._obj
 

examples/field_evolution.py

@@ -32,9 +32,9 @@
 pivot_cube = eds.datacube(
     "sentinel-2-l2a",
     intersects=pivot,
-    datetime=["2022-07-01", "2022-08-31"],
+    datetime=["2023-07-01","2023-07-10"],
     assets=["red", "green", "blue"],
-    mask_with="ag_cloud_mask",
+    mask_with="native",
     clear_cover=50,
 )