ssb-sgis 1.1.5__tar.gz → 1.1.7__tar.gz

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: ssb-sgis
-Version: 1.1.5
+Version: 1.1.7
 Summary: GIS functions used at Statistics Norway.
 Home-page: https://github.com/statisticsnorway/ssb-sgis
 License: MIT

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "ssb-sgis"
-version = "1.1.5"
+version = "1.1.7"
 description = "GIS functions used at Statistics Norway."
 authors = ["Morten Letnes <morten.letnes@ssb.no>"]
 license = "MIT"
@@ -56,6 +56,7 @@ bucket = ["dapla-toolbelt", "gcsfs"]
 torch = ["torch", "torchgeo"]
 xarray = ["xarray", "rioxarray"]
 test = ["xarray", "rioxarray", "torch", "torchgeo", "dask"]
+
 [tool.poetry.group.dev.dependencies]
 pygments = ">=2.10.0"
 black = { extras = ["jupyter"], version = ">=23.1.0" }

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/src/sgis/geopandas_tools/cleaning.py

@@ -101,18 +101,18 @@ def coverage_clean(
 
     _cleaning_checks(gdf, tolerance, duplicate_action)
 
-    if not gdf.index.is_unique:
-        gdf = gdf.reset_index(drop=True)
+    # if not gdf.index.is_unique:
+    #     gdf = gdf.reset_index(drop=True)
 
-    gdf = make_all_singlepart(gdf).loc[
-        lambda x: x.geom_type.isin(["Polygon", "MultiPolygon"])
-    ]
+    # gdf = make_all_singlepart(gdf).loc[
+    #     lambda x: x.geom_type.isin(["Polygon", "MultiPolygon"])
+    # ]
 
-    gdf = safe_simplify(gdf, PRECISION)
+    # gdf = safe_simplify(gdf, PRECISION)
 
     gdf = (
         clean_geoms(gdf)
-        .pipe(make_all_singlepart)
+        .pipe(make_all_singlepart, ignore_index=True)
         .loc[lambda x: x.geom_type.isin(["Polygon", "MultiPolygon"])]
     )
 
@@ -475,14 +475,14 @@ def _dissolve_thick_double_and_update(gdf, double, thin_double):
     large = (
         double.loc[~double["_double_idx"].isin(thin_double["_double_idx"])]
         .drop(columns="_double_idx")
-        # .pipe(sort_large_first)
-        .sort_values("_poly_idx")
+        .pipe(sort_small_first)
+        # .sort_values("_poly_idx")
         .pipe(update_geometries, geom_type="polygon")
     )
     return (
-        clean_overlay(gdf, large, how="update")
-        # .pipe(sort_large_first)
-        .sort_values("_poly_idx").pipe(update_geometries, geom_type="polygon")
+        clean_overlay(gdf, large, how="update").pipe(sort_small_first)
+        # .sort_values("_poly_idx")
+        .pipe(update_geometries, geom_type="polygon")
     )
 
 
@@ -534,7 +534,8 @@ def split_and_eliminate_by_longest(
     **kwargs,
 ) -> GeoDataFrame | tuple[GeoDataFrame]:
     if not len(to_eliminate):
-        return gdf
+        gdf = (gdf,) if isinstance(gdf, GeoDataFrame) else gdf
+        return (*gdf, to_eliminate)
 
     if not isinstance(gdf, (GeoDataFrame, GeoSeries)):
         as_gdf = pd.concat(gdf, ignore_index=True)
@@ -596,7 +597,10 @@ def split_by_neighbors(df, split_by, tolerance, grid_size=None) -> GeoDataFrame:
 
     intersecting_lines = (
         clean_overlay(
-            to_lines(split_by), buff(df, tolerance), how="identity", grid_size=grid_size
+            to_lines(split_by.explode(ignore_index=True)[lambda x: x.length > 0]),
+            buff(df, tolerance),
+            how="identity",
+            grid_size=grid_size,
         )
         .pipe(get_line_segments)
         .reset_index(drop=True)

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/src/sgis/geopandas_tools/duplicates.py

@@ -164,6 +164,8 @@ def update_geometries(
 
     copied.index = copied.index.map(index_mapper)
 
+    copied = make_all_singlepart(copied)
+
     # TODO check why polygons dissappear in rare cases. For now, just add back the missing
     dissapeared = sfilter_inverse(gdf, copied.buffer(-PRECISION))
     copied = pd.concat([copied, dissapeared])
@@ -307,7 +309,7 @@ def _get_intersecting_geometries(
     right["idx_right"] = right.index
 
     left = (
-        gdf
+        gdf.copy()
         if not any("index_" in str(col) for col in gdf)
         else gdf.loc[:, lambda x: x.columns.difference({"index_right", "index_left"})]
     )

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/src/sgis/geopandas_tools/general.py

@@ -594,8 +594,9 @@ def to_lines(
     """
     gdf = (
         pd.concat(df.assign(**{"_df_idx": i}) for i, df in enumerate(gdfs))
-        .pipe(make_all_singlepart, ignore_index=True)
+        .pipe(make_all_singlepart)
         .pipe(clean_geoms)
+        .pipe(make_all_singlepart, ignore_index=True)
     )
     geom_col = gdf.geometry.name
 
@@ -607,7 +608,9 @@ def to_lines(
     if (geoms.geom_type == "Polygon").all():
         geoms = polygons_to_lines(geoms, copy=copy)
     elif (geoms.geom_type != "LineString").any():
-        raise ValueError("Point geometries not allowed in 'to_lines'.")
+        raise ValueError(
+            f"Point geometries not allowed in 'to_lines'. {geoms.geom_type.value_counts()}"
+        )
 
     gdf.geometry.loc[:] = geoms
 

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/src/sgis/io/dapla_functions.py

@@ -86,6 +86,9 @@ def read_geopandas(
     """
     file_system = _get_file_system(file_system, kwargs)
 
+    if isinstance(gcs_path, (Path | os.PathLike)):
+        gcs_path = str(gcs_path)
+
     if not isinstance(gcs_path, (str | Path | os.PathLike)):
         return _read_geopandas_from_iterable(
             gcs_path,
@@ -96,12 +99,14 @@ def read_geopandas(
             **kwargs,
         )
 
-    if (
+    single_eq_filter = (
         isinstance(filters, Iterable)
         and len(filters) == 1
         and ("=" in next(iter(filters)) or "==" in next(iter(filters)))
-    ):
-        # try to read only files in the relevant partition, because glob is slow without GCSFileSystem
+    )
+    # try to read files in subfolder path / "column=value"
+    # because glob is slow without GCSFileSystem from the root partition
+    if single_eq_filter:
         try:
             expression = "".join(next(iter(filters))).replace("==", "=")
             glob_func = _get_glob(file_system)
@@ -664,7 +669,7 @@ def _read_geopandas(
         return read_func(file, **kwargs)
     except ValueError as e:
         if "Missing geo metadata" not in str(e) and "geometry" not in str(e):
-            raise e
+            raise e.__class__(f"{e.__class__.__name__}: {e} for {file}. ") from e
         df = getattr(pd, f"read_{file_format}")(file, **kwargs)
         if not len(df):
             return GeoDataFrame(df)

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/src/sgis/maps/explore.py

@@ -441,11 +441,10 @@ class Explore(Map):
             self.cmap_start = self.kwargs.pop("cmap_start", 0)
             self.cmap_stop = self.kwargs.pop("cmap_stop", 256)
 
-        # if self._gdf.crs is None:
-        #     self.kwargs["crs"] = "Simple"
-
         self.original_crs = self.gdf.crs
 
+        self._to_categorical()
+
     def __repr__(self) -> str:
         """Representation."""
         return f"{self.__class__.__name__}({len(self)})"

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/src/sgis/maps/httpserver.py

@@ -1,4 +1,5 @@
 import os
+import socket
 import webbrowser
 from http.server import BaseHTTPRequestHandler
 from http.server import HTTPServer
@@ -7,8 +8,19 @@ from IPython.display import HTML
 from IPython.display import display
 
 
+def find_available_port(start_port: int, max_attempts: int = 10) -> int:
+    """Find an available port starting from `start_port`."""
+    for port in range(start_port, start_port + max_attempts):
+        with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as sock:
+            if sock.connect_ex(("127.0.0.1", port)) != 0:
+                return port  # Port is available
+    raise RuntimeError("No available ports found in range.")
+
+
 def run_html_server(contents: str | None = None, port: int = 3000) -> None:
     """Run a simple, temporary http web server for serving static HTML content."""
+    port = find_available_port(port)
+
     if "JUPYTERHUB_SERVICE_PREFIX" in os.environ:
         # Create a link using the https://github.com/jupyterhub/jupyter-server-proxy
         display_address = os.environ["JUPYTERHUB_SERVICE_PREFIX"] + f"proxy/{port}/"

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/src/sgis/maps/map.py

@@ -110,6 +110,7 @@ class Map:
         nan_color="#c2c2c2",
         scheme: str = DEFAULT_SCHEME,
         cmap: str | None = None,
+        categorical: bool | None = None,
         **kwargs,
     ) -> None:
         """Initialiser.
@@ -124,8 +125,11 @@ class Map:
             scheme: Classification scheme to be used.
             cmap (str): Colormap of the plot. See:
                 https://matplotlib.org/stable/tutorials/colors/colormaps.html
+            categorical: Set to True to convert 'column' to string values.
             **kwargs: Arbitrary keyword arguments.
         """
+        self.kwargs = {}
+
         gdfs, column, kwargs = self._separate_args(gdfs, column, kwargs)
 
         self._column = column
@@ -173,7 +177,6 @@ class Map:
         self.labels = new_labels
 
         # pop all geometry-like items from kwargs into self._gdfs
-        self.kwargs = {}
         i = 0
         for key, value in kwargs.items():
             try:
@@ -198,9 +201,13 @@ class Map:
                 f"length as gdfs ({len(gdfs)}). Got len {len(show)}"
             )
 
+        if categorical is not None:
+            self._is_categorical = categorical
+
         if not self._gdfs or not any(len(gdf) for gdf in self._gdfs):
             self._gdfs = []
-            self._is_categorical = True
+            if categorical is None:
+                self._is_categorical = True
             self._unique_values = []
             self._nan_idx = []
             return
@@ -209,7 +216,8 @@ class Map:
             self._set_labels()
 
         self._gdfs = self._to_common_crs_and_one_geom_col(self._gdfs)
-        self._is_categorical = self._check_if_categorical()
+        if categorical is None:
+            self._is_categorical = self._check_if_categorical()
 
         if self._column:
             self._fillna_if_col_is_missing()
@@ -230,11 +238,34 @@ class Map:
             self._gdf = pd.concat(self._gdfs, ignore_index=True)
 
         self._nan_idx = self._gdf[self._column].isna()
+        self._to_categorical()
         self._get_unique_values()
 
-    def __getattr__(self, attr: str) -> Any:
-        """Search for attribute in kwargs."""
-        return self.kwargs.get(attr, super().__getattribute__(attr))
+    def _to_categorical(self):
+        if not (self._is_categorical and self.column is not None):
+            return
+
+        def to_string_via_int(series):
+            if not pd.api.types.is_numeric_dtype(series):
+                return series.astype("string")
+            try:
+                series = series.astype(float)
+            except ValueError:
+                return series
+            no_decimals: bool = (series.dropna() % 1 == 0).all()
+            if no_decimals:
+                return series.astype("Int64").astype("string")
+            else:
+                return series.astype("string")
+
+        for i, gdf in enumerate(self._gdfs):
+            if self.column in gdf:
+                self._gdfs[i][self.column] = to_string_via_int(gdf[self.column])
+        self._gdf[self.column] = to_string_via_int(self._gdf[self.column])
+
+    # def __getattr__(self, attr: str) -> Any:
+    #     """Search for attribute in kwargs."""
+    #     return self.kwargs.get(attr, super().__getattribute__(attr))
 
     def __bool__(self) -> bool:
         """True of any gdfs with more than 0 rows."""
@@ -524,12 +555,6 @@ class Map:
         )
         n = n + maybe_area + maybe_length
 
-        if n == 0:
-            raise ValueError(
-                f"The column {self._column!r} is not present in any "
-                "of the passed GeoDataFrames."
-            )
-
     def _check_if_categorical(self) -> bool:
         """Quite messy this..."""
         if not self._column or not self._gdfs:

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/src/sgis/parallel/parallel.py

@@ -1117,7 +1117,6 @@ def chunkwise(
         iterable_chunked: list[Iterable] = [
             to_type(chunk) for chunk in np.array_split(list(iterable), n_chunks)
         ]
-
     return Parallel(processes, backend=backend).map(
         func,
         iterable_chunked,

{ssb_sgis-1.1.5 → ssb_sgis-1.1.7}/src/sgis/raster/image_collection.py

@@ -8,6 +8,7 @@ import re
 import time
 from abc import abstractmethod
 from collections.abc import Callable
+from collections.abc import Generator
 from collections.abc import Iterable
 from collections.abc import Iterator
 from collections.abc import Sequence
@@ -86,6 +87,14 @@ except ImportError:
         raise ImportError("xarray")
 
 
+try:
+    from gcsfs.core import GCSFile
+except ImportError:
+
+    class GCSFile:
+        """Placeholder."""
+
+
 from ..geopandas_tools.bounds import get_total_bounds
 from ..geopandas_tools.conversion import to_bbox
 from ..geopandas_tools.conversion import to_gdf
@@ -2645,9 +2654,10 @@ class ImageCollection(_ImageBase):
 
         other = to_shapely(other)
 
-        intersects_list: pd.Series = GeoSeries(
-            [img.union_all() for img in self]
-        ).intersects(other)
+        with ThreadPoolExecutor() as executor:
+            bounds_iterable: Generator[Polygon] = executor.map(_union_all, self)
+
+        intersects_list: pd.Series = GeoSeries(list(bounds_iterable)).intersects(other)
 
         self.images = [
             image
@@ -3479,6 +3489,10 @@ def _open_raster(path: str | Path) -> rasterio.io.DatasetReader:
         return rasterio.open(file)
 
 
+def _union_all(obj: _ImageBase) -> Polygon:
+    return obj.union_all()
+
+
 def _read_mask_array(self: Band | Image, **kwargs) -> np.ndarray:
     mask_band_id = self.masking["band_id"]
     mask_paths = [path for path in self._all_file_paths if mask_band_id in path]