ENH: Add from_wkt and from_wkb tools (#293)

Co-authored-by: Joris Van den Bossche <[email protected]>

Author: giorgiobasile

dask_geopandas/__init__.py

@@ -5,6 +5,8 @@
 if backends.QUERY_PLANNING_ON:
     from .expr import (
         points_from_xy,
+        from_wkt,
+        from_wkb,
         GeoDataFrame,
         GeoSeries,
         from_geopandas,
@@ -13,6 +15,8 @@
 else:
     from .core import (
         points_from_xy,
+        from_wkt,
+        from_wkb,
         GeoDataFrame,
         GeoSeries,
         from_geopandas,
@@ -30,6 +34,8 @@
 
 __all__ = [
     "points_from_xy",
+    "from_wkt",
+    "from_wkb",
     "GeoDataFrame",
     "GeoSeries",
     "from_geopandas",

dask_geopandas/core.py

@@ -882,6 +882,56 @@ def func(data, x, y, z):
     )
 
 
+def from_wkt(wkt, crs=None):
+    """
+    Convert dask.dataframe.Series of WKT objects to a GeoSeries.
+
+    Parameters
+    ----------
+    wkt: dask Series
+        A dask Series containing WKT objects.
+    crs: value, optional
+            Coordinate Reference System of the geometry objects. Can be anything
+            accepted by
+            :meth:`pyproj.CRS.from_user_input() <pyproj.crs.CRS.from_user_input>`,
+            such as an authority string (eg "EPSG:4326") or a WKT string.
+
+    Returns
+    -------
+    GeoSeries
+    """
+
+    def func(data):
+        return geopandas.GeoSeries.from_wkt(data, index=data.index, crs=crs)
+
+    return wkt.map_partitions(func, meta=geopandas.GeoSeries(), token="from_wkt")
+
+
+def from_wkb(wkb, crs=None):
+    """
+    Convert dask.dataframe.Series of WKB objects to a GeoSeries.
+
+    Parameters
+    ----------
+    wkb: dask Series
+        A dask Series containing WKB objects.
+    crs: value, optional
+            Coordinate Reference System of the geometry objects. Can be anything
+            accepted by
+            :meth:`pyproj.CRS.from_user_input() <pyproj.crs.CRS.from_user_input>`,
+            such as an authority string (eg "EPSG:4326") or a WKT string.
+
+    Returns
+    -------
+    GeoSeries
+    """
+
+    def func(data):
+        return geopandas.GeoSeries.from_wkb(data, index=data.index, crs=crs)
+
+    return wkb.map_partitions(func, meta=geopandas.GeoSeries(), token="from_wkb")
+
+
 for name in [
     "area",
     "geom_type",

dask_geopandas/expr.py

@@ -914,6 +914,56 @@ def func(data, x, y, z):
     )
 
 
+def from_wkt(wkt, crs=None):
+    """
+    Convert dask.dataframe.Series of WKT objects to a GeoSeries.
+
+    Parameters
+    ----------
+    wkt: dask Series
+        A dask Series containing WKT objects.
+    crs: value, optional
+            Coordinate Reference System of the geometry objects. Can be anything
+            accepted by
+            :meth:`pyproj.CRS.from_user_input() <pyproj.crs.CRS.from_user_input>`,
+            such as an authority string (eg "EPSG:4326") or a WKT string.
+
+    Returns
+    -------
+    GeoSeries
+    """
+
+    def func(data):
+        return geopandas.GeoSeries.from_wkt(data, index=data.index, crs=crs)
+
+    return wkt.map_partitions(func, meta=geopandas.GeoSeries(), token="from_wkt")
+
+
+def from_wkb(wkb, crs=None):
+    """
+    Convert dask.dataframe.Series of WKB objects to a GeoSeries.
+
+    Parameters
+    ----------
+    wkb: dask Series
+        A dask Series containing WKB objects.
+    crs: value, optional
+            Coordinate Reference System of the geometry objects. Can be anything
+            accepted by
+            :meth:`pyproj.CRS.from_user_input() <pyproj.crs.CRS.from_user_input>`,
+            such as an authority string (eg "EPSG:4326") or a WKT string.
+
+    Returns
+    -------
+    GeoSeries
+    """
+
+    def func(data):
+        return geopandas.GeoSeries.from_wkb(data, index=data.index, crs=crs)
+
+    return wkb.map_partitions(func, meta=geopandas.GeoSeries(), token="from_wkb")
+
+
 for name in [
     "area",
     "geom_type",

dask_geopandas/tests/test_core.py

@@ -183,6 +183,32 @@ def test_points_from_xy_with_crs():
     assert_geoseries_equal(actual.compute(), expected)
 
 
+def test_from_wkt():
+    wkt = [
+        "POLYGON ((-64.8 32.3, -65.5 18.3, -80.3 25.2, -64.8 32.3))",
+        "POLYGON ((-81.079102 35.496456, -81.166992 31.914868, -75.541992 31.914868, -75.629883 35.675147, -81.079102 35.496456))",  # noqa E501
+    ]
+    expected = geopandas.GeoSeries.from_wkt(wkt, crs="EPSG:4326")
+    df = pd.DataFrame({"wkt": wkt})
+    ddf = dd.from_pandas(df, npartitions=2)
+    actual = dask_geopandas.from_wkt(ddf["wkt"], crs="EPSG:4326")
+    assert isinstance(actual, dask_geopandas.GeoSeries)
+    assert_geoseries_equal(actual.compute(), expected)
+
+
+def test_from_wkb():
+    wkb = [
+        "0103000000010000000400000033333333333350c0666666666626404000000000006050c0cdcccccccc4c324033333333331354c0333333333333394033333333333350c06666666666264040",
+        "0103000000010000000500000016c3d501104554c095f3c5de8bbf414064ac36ffaf4a54c02c280cca34ea3f4064ac36ffafe252c02c280cca34ea3f409c53c90050e852c00b7f86376bd6414016c3d501104554c095f3c5de8bbf4140",
+    ]
+    expected = geopandas.GeoSeries.from_wkb(wkb, crs="EPSG:4326")
+    df = pd.DataFrame({"wkb": wkb})
+    ddf = dd.from_pandas(df, npartitions=2)
+    actual = dask_geopandas.from_wkb(ddf["wkb"], crs="EPSG:4326")
+    assert isinstance(actual, dask_geopandas.GeoSeries)
+    assert_geoseries_equal(actual.compute(), expected)
+
+
 def test_geodataframe_crs(geodf_points_crs):
     df = geodf_points_crs
     original = df.crs

doc/source/docs/reference/tools.rst

@@ -9,3 +9,5 @@ Tools
    sjoin
    clip
    points_from_xy
+   from_wkt
+   from_wkb