feat(GeoDataFrame support): add in geopandas GeoDataFrame support

autotest/test_binaryfile.py

@@ -275,7 +275,7 @@ def test_load_binary_head_file(example_data_path):
 
 def test_plot_binary_head_file(example_data_path):
     hf = HeadFile(example_data_path / "freyberg" / "freyberg.githds")
-    hf.mg.set_coord_info(xoff=1000.0, yoff=200.0, angrot=15.0)
+    hf.modelgrid.set_coord_info(xoff=1000.0, yoff=200.0, angrot=15.0)
 
     assert isinstance(hf.plot(), Axes)
     plt.close()

autotest/test_formattedfile.py

@@ -69,7 +69,7 @@ def test_headfile_build_index(example_data_path):
 def test_formattedfile_reference(example_data_path):
     h = FormattedHeadFile(example_data_path / "mf2005_test" / "test1tr.githds")
     assert isinstance(h, FormattedHeadFile)
-    h.mg.set_coord_info(xoff=1000.0, yoff=200.0, angrot=15.0)
+    h.modelgrid.set_coord_info(xoff=1000.0, yoff=200.0, angrot=15.0)
 
     assert isinstance(h.plot(masked_values=[6999.000]), Axes)
     plt.close()

autotest/test_modpathfile.py

@@ -306,11 +306,10 @@ def test_get_destination_endpoint_data(
     )
 
 
-@pytest.mark.parametrize("longfieldname", [True, False])
 @requires_exe("mf6", "mp7")
-@requires_pkg("pyshp", "shapely", name_map={"pyshp": "shapefile"})
-def test_write_shapefile(function_tmpdir, mp7_small, longfieldname):
-    from shapefile import Reader
+@requires_pkg("geopandas", "shapely")
+def test_write_shapefile(function_tmpdir, mp7_small):
+    import geopandas as gpd
 
     # setup and run model, then copy outputs to function_tmpdir
     sim, forward_model_name, _, _, _ = mp7_small
@@ -330,13 +329,7 @@ def test_write_shapefile(function_tmpdir, mp7_small, longfieldname):
     # define shapefile path
     shp_file = ws / "pathlines.shp"
 
-    # add a column to the pathline recarray
-    fieldname = "newfield" + ("longname" if longfieldname else "")
-    fieldval = "x"
-    pathlines = [
-        rfn.append_fields(pl, fieldname, list(repeat(fieldval, len(pl))), dtypes="|S1")
-        for pl in pathlines
-    ]
+    pline_names = [name[:10] for name in pathlines[0].dtype.names]
 
     # write the pathline recarray to shapefile
     pathline_file.write_shapefile(
@@ -350,8 +343,7 @@ def test_write_shapefile(function_tmpdir, mp7_small, longfieldname):
     assert shp_file.is_file()
 
     # load shapefile
-    with Reader(shp_file) as reader:
-        fieldnames = [f[0] for f in reader.fields[1:]]
-        fieldname = "newfiname_" if longfieldname else fieldname
-        assert fieldname in fieldnames
-        assert all(r[fieldname] == fieldval for r in reader.iterRecords())
+    gdf = gpd.read_file(shp_file)
+    fieldnames = list(gdf)
+    for fname in pline_names:
+        assert fname in fieldnames

flopy/mbase.py

@@ -759,6 +759,39 @@ def _output_msg(self, i, add=True):
             f"{txt2} the output list."
         )
 
+    def to_geo_dataframe(self, gdf=None, kper=0):
+        """
+        Method to build a Geodataframe from model inputs. Note: transient data
+        will only be exported for a single stress period.
+
+        Parameters
+        ----------
+        gdf : GeoDataFrame
+            optional geopandas geodataframe object to add data to. Default is None
+        kper : int
+            stress period to get transient data from
+
+        Returns
+        -------
+            gdf : GeoDataFrame
+        """
+        if gdf is None:
+            modelgrid = self.modelgrid
+            if modelgrid is not None:
+                gdf = modelgrid.geo_dataframe
+            else:
+                raise AttributeError(
+                    "model does not have a grid instance, please supply a geodataframe"
+                )
+
+        for package in self.packagelist:
+            if package.package_type in ("hfb6",):
+                continue
+            if callable(getattr(package, "to_geo_dataframe", None)):
+                gdf = package.to_geo_dataframe(gdf, kper=kper, sparse=False)
+
+        return gdf
+
     def add_output_file(
         self,
         unit,

flopy/mf6/data/mfdataarray.py

@@ -323,6 +323,64 @@ def data(self):
         """Returns array data.  Calls get_data with default parameters."""
         return self._get_data()
 
+    def to_geo_dataframe(self, gdf=None, name=None, forgive=False, **kwargs):
+        """
+        Method to add an input array to a geopandas GeoDataFrame
+
+        Parameters
+        ----------
+        gdf : GeoDataFrame
+            optional GeoDataFrame object
+        name : str
+            optional attribute name, default uses util2d name
+        forgive : bool
+            optional flag to continue if data shape not compatible with GeoDataFrame
+
+        Returns
+        -------
+            geopandas GeoDataFrame
+        """
+        if self.model is None:
+            return gdf
+        else:
+            modelgrid = self.model.modelgrid
+            if gdf is None:
+                if modelgrid is None:
+                    return gdf
+                gdf = modelgrid.geo_dataframe
+
+            if modelgrid is not None:
+                if modelgrid.grid_type != "unstructured":
+                    ncpl = modelgrid.ncpl
+                else:
+                    ncpl = modelgrid.nnodes
+            else:
+                ncpl = len(gdf)
+
+            if name is None:
+                name = self.name
+
+            data = self.array
+            if data is None:
+                return gdf
+
+            if data.size == ncpl:
+                gdf[name] = data.ravel()
+
+            elif data.size % ncpl == 0:
+                data = data.reshape((-1, ncpl))
+                for ix, arr in enumerate(data):
+                    aname = f"{name}_{ix}"
+                    gdf[aname] = arr
+            elif forgive:
+                return gdf
+            else:
+                raise ValueError(
+                    f"Data size {data.size} not compatible with dataframe length {ncpl}"
+                )
+
+            return gdf
+
     def new_simulation(self, sim_data):
         """Initialize MFArray object for a new simulation
 
@@ -1890,6 +1948,50 @@ def _build_period_data(
                     output[sp] = data
         return output
 
+    def to_geo_dataframe(self, gdf=None, kper=0, forgive=False, **kwargs):
+        """
+
+        """
+        if self.model is None:
+            return gdf
+        else:
+            modelgrid = self.model.modelgrid
+            if gdf is None:
+                if modelgrid is None:
+                    return gdf
+                gdf = modelgrid.geo_dataframe
+
+            if modelgrid is not None:
+                if modelgrid.grid_type != "unstructured":
+                    ncpl = modelgrid.ncpl
+                else:
+                    ncpl = modelgrid.nnodes
+            else:
+                ncpl = len(gdf)
+
+            if self.array is None:
+                return gdf
+
+            name = f"{self.path[1]}_{self.name}"
+
+            data = self.get_data(key=kper, apply_mult=True)
+            if data.size == ncpl:
+                gdf[name] = data.ravel()
+
+            elif data.size % ncpl == 0:
+                data = data.reshape((-1, ncpl))
+                for ix, arr in enumerate(data):
+                    aname = f"{name}_{ix}"
+                    gdf[aname] = arr
+            elif forgive:
+                return gdf
+            else:
+                raise ValueError(
+                    f"Data size {data.size} not compatible with dataframe length {ncpl}"
+                )
+
+            return gdf
+
     def set_record(self, data_record):
         """Sets data and metadata at layer `layer` and time `key` to
         `data_record`.  For unlayered data do not pass in `layer`.

flopy/mf6/data/mfdatalist.py

@@ -145,6 +145,55 @@ def to_array(self, kper=0, mask=False):
         model_grid = self.data_dimensions.get_model_grid()
         return list_to_array(sarr, model_grid, kper, mask)
 
+    def to_geo_dataframe(self, gdf=None, sparse=False, **kwargs):
+        """
+        Method to add data to a GeoDataFrame for exporting as a geospatial file
+
+        Parameters
+        ----------
+        gdf : GeoDataFrame
+            optional GeoDataFrame instance. If GeoDataFrame is None, one will be
+            constructed from modelgrid information
+        sparse : bool
+            boolean flag for sparse dataframe construction. Default is False
+
+        Returns
+        -------
+            GeoDataFrame
+        """
+        if self.model is None:
+            return gdf
+        else:
+            modelgrid = self.model.modelgrid
+            if modelgrid is None:
+                return gdf
+
+            if gdf is None:
+                gdf = modelgrid.geo_dataframe
+
+            data = self.to_array(mask=True)
+            if data is None:
+                return gdf
+
+            col_names = []
+            for name, array3d in data.items():
+                aname = f"{self.path[1].lower()}_{name}"
+                if modelgrid.grid_type == "unstructured":
+                    array = array3d.ravel()
+                    gdf[aname] = array
+                    col_names.append(aname)
+                else:
+                    for lay in range(modelgrid.nlay):
+                        arr = array3d[lay].ravel()
+                        gdf[f"{aname}_{lay}"] = arr.ravel()
+                        col_names.append(f"{aname}_{lay}")
+
+            if sparse:
+                gdf = gdf.dropna(subset=col_names, how="all")
+                gdf = gdf.dropna(axis="columns", how="all")
+
+            return gdf
+
     def new_simulation(self, sim_data):
         """Initialize MFList object for a new simulation.
 
@@ -1596,6 +1645,57 @@ def to_array(self, kper=0, mask=False):
         """Returns list data as an array."""
         return super().to_array(kper, mask)
 
+    def to_geo_dataframe(self, gdf=None, kper=0, sparse=False, **kwargs):
+        """
+        Method to add data to a GeoDataFrame for exporting as a geospatial file
+
+        Parameters
+        ----------
+        gdf : GeoDataFrame
+            optional GeoDataFrame instance. If GeoDataFrame is None, one will be
+            constructed from modelgrid information
+        kper : int
+            stress period to export
+        sparse : bool
+            boolean flag for sparse dataframe construction. Default is False
+
+        Returns
+        -------
+            GeoDataFrame
+        """
+        if self.model is None:
+            return gdf
+        else:
+            modelgrid = self.model.modelgrid
+            if modelgrid is None:
+                return gdf
+
+            if gdf is None:
+                gdf = modelgrid.geo_dataframe
+
+            data = self.to_array(kper=kper, mask=True)
+            if data is None:
+                return gdf
+
+            col_names = []
+            for name, array3d in data.items():
+                aname = f"{self.path[1].lower()}_{name}"
+                if modelgrid.grid_type == "unstructured":
+                    array = array3d.ravel()
+                    gdf[aname] = array
+                    col_names.append(aname)
+                else:
+                    for lay in range(modelgrid.nlay):
+                        arr = array3d[lay].ravel()
+                        gdf[f"{aname}_{lay}"] = arr.ravel()
+                        col_names.append(f"{aname}_{lay}")
+
+            if sparse:
+                gdf = gdf.dropna(subset=col_names, how="all")
+                gdf = gdf.dropna(axis="columns", how="all")
+
+            return gdf
+
     def remove_transient_key(self, transient_key):
         """Remove transient stress period key.  Method is used
         internally by FloPy and is not intended to the end user.