ISEA4TPandas module

ISEA4TPandas

Source code in vgridpandas/isea4tpandas.py

@pd.api.extensions.register_dataframe_accessor("isea4t")
class ISEA4TPandas:
    def __init__(self, df: DataFrame):
        self._df = df

    def latlon2isea4t(
        self,
        resolution: int,
        lat_col: str = "lat",
        lon_col: str = "lon",
        set_index: bool = False,
    ) -> AnyDataFrame:
        """Adds ISEA4T ID to (Geo)DataFrame.

        pd.DataFrame: uses `lat_col` and `lon_col` (default `lat` and `lon`)
        gpd.GeoDataFrame: uses `geometry`

        Assumes coordinates in epsg=4326.

        Parameters
        ----------
        resolution : int
            ISEA4T resolution
        lat_col : str
            Name of the latitude column (if used), default 'lat'
        lon_col : str
            Name of the longitude column (if used), default 'lon'
        set_index : bool
            If True, the columns with ISEA4T ID is set as index, default 'True'

        Returns
        -------
        (Geo)DataFrame with ISEA4T IDs added
        """
        if isinstance(self._df, gpd.GeoDataFrame):
            lons = self._df.geometry.x
            lats = self._df.geometry.y
        else:
            lons = self._df[lon_col]
            lats = self._df[lat_col]

        isea4t_ids = [
            latlon_to_isea4t(lat, lon, resolution) for lat, lon in zip(lats, lons)
        ]

        isea4t_col = ISEA4T_COL
        assign_arg = {isea4t_col: isea4t_ids, f"{isea4t_col}_res": resolution}
        df = self._df.assign(**assign_arg)
        if set_index:
            return df.set_index(isea4t_col)         
        return df

    def isea4t2geo(
        self, isea4t_col: str = None, fix_antimeridian: Optional[str] = None
    ) -> GeoDataFrame:
        """Add geometry with ISEA4T geometry to the DataFrame."""
        if isea4t_col is not None:
            if isea4t_col not in self._df.columns:  
                raise ValueError(f"Column '{isea4t_col}' not found in DataFrame")
            ids = self._df[isea4t_col]
        else:       
            if ISEA4T_COL not in self._df.columns:
                raise ValueError(f"Column '{ISEA4T_COL}' not found in DataFrame")
            ids = self._df[ISEA4T_COL]
        return dggs_ids_to_geodataframe(
            self._df, ids, isea4t_to_geo, fix_antimeridian=fix_antimeridian
        )   

    def polyfill(
        self,
        resolution: int,
        predicate: str = None,
        compact: bool = False,
        explode: bool = False,
        fix_antimeridian: Optional[str] = None,
    ) -> AnyDataFrame:
        """
        Parameters
        ----------
        resolution : int
            isea4t resolution
        predicate : str, optional
            Spatial predicate to apply ('intersect', 'within', 'centroid_within', 'largest_overlap')
        compact : bool, optional
            Whether to compact the isea4t IDs
        explode : bool
            If True, will explode the resulting list vertically.
            All other columns' values are copied.
            Default: False
        fix_antimeridian : str, optional
            Antimeridian fixing method passed to ``isea4t2geo``
        """

        result = self._df.geometry.apply(
            lambda geom: polyfill_row(
                geom, resolution, predicate, compact, fix_antimeridian
            )
        )

        if not explode:
            assign_args = {ISEA4T_COL: result}
            return self._df.assign(**assign_args)

        result = result.explode().to_frame(ISEA4T_COL)
        return self._df.join(result)

    def isea4tbin(
        self,
        resolution: int,
        stats: str = "count",
        numeric_col: str = None,
        category_col: str = None,
        lat_col: str = "lat",
        lon_col: str = "lon",
        fix_antimeridian: Optional[str] = None,
    ) -> GeoDataFrame:
        """
        Bin points into isea4t cells and compute statistics.
        """
        isea4t_col = ISEA4T_COL
        df = self.latlon2isea4t(resolution, lat_col, lon_col)
        result = aggregate_bin(df, isea4t_col, stats, numeric_col, category_col)
        return result.isea4t.isea4t2geo(
            isea4t_col=isea4t_col, fix_antimeridian=fix_antimeridian
        )

isea4t2geo(isea4t_col=None, fix_antimeridian=None)

Add geometry with ISEA4T geometry to the DataFrame.

Source code in vgridpandas/isea4tpandas.py

def isea4t2geo(
    self, isea4t_col: str = None, fix_antimeridian: Optional[str] = None
) -> GeoDataFrame:
    """Add geometry with ISEA4T geometry to the DataFrame."""
    if isea4t_col is not None:
        if isea4t_col not in self._df.columns:  
            raise ValueError(f"Column '{isea4t_col}' not found in DataFrame")
        ids = self._df[isea4t_col]
    else:       
        if ISEA4T_COL not in self._df.columns:
            raise ValueError(f"Column '{ISEA4T_COL}' not found in DataFrame")
        ids = self._df[ISEA4T_COL]
    return dggs_ids_to_geodataframe(
        self._df, ids, isea4t_to_geo, fix_antimeridian=fix_antimeridian
    )   

isea4tbin(resolution, stats='count', numeric_col=None, category_col=None, lat_col='lat', lon_col='lon', fix_antimeridian=None)

Bin points into isea4t cells and compute statistics.

Source code in vgridpandas/isea4tpandas.py

def isea4tbin(
    self,
    resolution: int,
    stats: str = "count",
    numeric_col: str = None,
    category_col: str = None,
    lat_col: str = "lat",
    lon_col: str = "lon",
    fix_antimeridian: Optional[str] = None,
) -> GeoDataFrame:
    """
    Bin points into isea4t cells and compute statistics.
    """
    isea4t_col = ISEA4T_COL
    df = self.latlon2isea4t(resolution, lat_col, lon_col)
    result = aggregate_bin(df, isea4t_col, stats, numeric_col, category_col)
    return result.isea4t.isea4t2geo(
        isea4t_col=isea4t_col, fix_antimeridian=fix_antimeridian
    )

latlon2isea4t(resolution, lat_col='lat', lon_col='lon', set_index=False)

Adds ISEA4T ID to (Geo)DataFrame.

pd.DataFrame: uses lat_col and lon_col (default lat and lon) gpd.GeoDataFrame: uses geometry

Assumes coordinates in epsg=4326.

Parameters

resolution : int ISEA4T resolution lat_col : str Name of the latitude column (if used), default 'lat' lon_col : str Name of the longitude column (if used), default 'lon' set_index : bool If True, the columns with ISEA4T ID is set as index, default 'True'

Returns

(Geo)DataFrame with ISEA4T IDs added

Source code in vgridpandas/isea4tpandas.py

def latlon2isea4t(
    self,
    resolution: int,
    lat_col: str = "lat",
    lon_col: str = "lon",
    set_index: bool = False,
) -> AnyDataFrame:
    """Adds ISEA4T ID to (Geo)DataFrame.

    pd.DataFrame: uses `lat_col` and `lon_col` (default `lat` and `lon`)
    gpd.GeoDataFrame: uses `geometry`

    Assumes coordinates in epsg=4326.

    Parameters
    ----------
    resolution : int
        ISEA4T resolution
    lat_col : str
        Name of the latitude column (if used), default 'lat'
    lon_col : str
        Name of the longitude column (if used), default 'lon'
    set_index : bool
        If True, the columns with ISEA4T ID is set as index, default 'True'

    Returns
    -------
    (Geo)DataFrame with ISEA4T IDs added
    """
    if isinstance(self._df, gpd.GeoDataFrame):
        lons = self._df.geometry.x
        lats = self._df.geometry.y
    else:
        lons = self._df[lon_col]
        lats = self._df[lat_col]

    isea4t_ids = [
        latlon_to_isea4t(lat, lon, resolution) for lat, lon in zip(lats, lons)
    ]

    isea4t_col = ISEA4T_COL
    assign_arg = {isea4t_col: isea4t_ids, f"{isea4t_col}_res": resolution}
    df = self._df.assign(**assign_arg)
    if set_index:
        return df.set_index(isea4t_col)         
    return df

polyfill(resolution, predicate=None, compact=False, explode=False, fix_antimeridian=None)

Parameters

resolution : int isea4t resolution predicate : str, optional Spatial predicate to apply ('intersect', 'within', 'centroid_within', 'largest_overlap') compact : bool, optional Whether to compact the isea4t IDs explode : bool If True, will explode the resulting list vertically. All other columns' values are copied. Default: False fix_antimeridian : str, optional Antimeridian fixing method passed to isea4t2geo

Source code in vgridpandas/isea4tpandas.py

def polyfill(
    self,
    resolution: int,
    predicate: str = None,
    compact: bool = False,
    explode: bool = False,
    fix_antimeridian: Optional[str] = None,
) -> AnyDataFrame:
    """
    Parameters
    ----------
    resolution : int
        isea4t resolution
    predicate : str, optional
        Spatial predicate to apply ('intersect', 'within', 'centroid_within', 'largest_overlap')
    compact : bool, optional
        Whether to compact the isea4t IDs
    explode : bool
        If True, will explode the resulting list vertically.
        All other columns' values are copied.
        Default: False
    fix_antimeridian : str, optional
        Antimeridian fixing method passed to ``isea4t2geo``
    """

    result = self._df.geometry.apply(
        lambda geom: polyfill_row(
            geom, resolution, predicate, compact, fix_antimeridian
        )
    )

    if not explode:
        assign_args = {ISEA4T_COL: result}
        return self._df.assign(**assign_args)

    result = result.explode().to_frame(ISEA4T_COL)
    return self._df.join(result)

poly2isea4t(geometry, resolution, predicate=None, compact=False, fix_antimeridian=None)

Convert polygon or line geometries to ISEA4T grid cells.

Mirrors polygon2isea4t and polyline2isea4t in vgrid (Windows only). Polygons are filtered with predicate; lines use intersection. Compact mode applies to polygons after predicate filtering.

Parameters:

Name	Type	Description	Default
resolution	int	ISEA4T resolution level [0..28]	required
geometry		Polygon, MultiPolygon, LineString, or MultiLineString	required
predicate	str	Spatial predicate for polygons ('intersect', 'within', 'centroid_within', 'largest_overlap')	None
compact	bool	Enable ISEA4T compact mode for polygons	False
fix_antimeridian	str	Antimeridian fixing method passed to isea4t2geo: shift, shift_balanced, shift_west, shift_east, split, none	None

Returns:

Name	Type	Description
list	list	List of ISEA4T cell ids

Source code in vgridpandas/isea4tpandas.py

def poly2isea4t(
    geometry,
    resolution: int,
    predicate: str = None,
    compact: bool = False,
    fix_antimeridian: Optional[str] = None,
) -> list:
    """
    Convert polygon or line geometries to ISEA4T grid cells.

    Mirrors ``polygon2isea4t`` and ``polyline2isea4t`` in vgrid (Windows only).
    Polygons are filtered with ``predicate``; lines use intersection.
    Compact mode applies to polygons after predicate filtering.

    Args:
        resolution (int): ISEA4T resolution level [0..28]
        geometry: Polygon, MultiPolygon, LineString, or MultiLineString
        predicate (str, optional): Spatial predicate for polygons
            ('intersect', 'within', 'centroid_within', 'largest_overlap')
        compact (bool, optional): Enable ISEA4T compact mode for polygons
        fix_antimeridian (str, optional): Antimeridian fixing method passed to
            ``isea4t2geo``: shift, shift_balanced, shift_west, shift_east, split, none

    Returns:
        list: List of ISEA4T cell ids
    """
    if platform.system() != "Windows":
        return []

    resolution = validate_isea4t_resolution(resolution)
    isea4t_ids = []
    if isinstance(geometry, (Polygon, LineString)):
        polys = [geometry]
    elif isinstance(geometry, (MultiPolygon, MultiLineString)):
        polys = list(geometry.geoms)
    else:
        return []

    for poly in polys:
        if poly is None or poly.is_empty:   
            continue

        is_line = isinstance(poly, LineString)
        bounding_child_cells = _isea4t_children_for_bounds(poly.bounds, resolution)

        poly_ids = []
        for child in bounding_child_cells:
            isea4t_id = child
            cell_polygon = isea4t_to_geo(
                isea4t_id, fix_antimeridian=fix_antimeridian
            )
            if is_line:
                if not cell_polygon.intersects(poly):
                    continue
            elif not check_predicate(cell_polygon, poly, predicate):
                continue
            poly_ids.append(isea4t_id)

        if compact and poly_ids and not is_line:
            poly_ids = list(isea4t_compact(poly_ids))

        isea4t_ids.extend(poly_ids)

    return list(dict.fromkeys(isea4t_ids))

polyfill_row(geometry, resolution, predicate=None, compact=False, fix_antimeridian=None)

Return cell ids covering a single row geometry.

Source code in vgridpandas/isea4tpandas.py

def polyfill_row(
    geometry,
    resolution,
    predicate=None,
    compact=False,
    fix_antimeridian: Optional[str] = None,
) -> list:
    """Return cell ids covering a single row geometry."""
    if isinstance(geometry, (Polygon, MultiPolygon)):
        tokens = set(
            poly2isea4t(geometry, resolution, predicate, compact, fix_antimeridian)
        )
    elif isinstance(geometry, (LineString, MultiLineString)):
        tokens = set(
            poly2isea4t(
                geometry,
                resolution,
                predicate="intersect",
                compact=False,
                fix_antimeridian=fix_antimeridian,
            )
        )
    else:
        raise TypeError(f"Unknown type {type(geometry)}")
    return list(tokens)