def clip_by_extent(gdf: gpd.geodataframe.GeoDataFrame,
bbox: List[Union[int, float]],
inplace: bool = False) -> gpd.geodataframe.GeoDataFrame:
"""
Clipping vector data by extent
Args:
gdf: GeoDataFrame to be clipped
bbox: list of bounds for the gdf to be clipped
inplace: - bool - default False -> copy of the current gdf is created
Return:
gdf: GeoDataFrame with the clipped values
"""
# Checking if the gdf is of type GeoDataFrame
if not isinstance(gdf, gpd.geodataframe.GeoDataFrame):
raise TypeError('gdf must be of type GeoDataFrame')
# Checking that the bbox is of type list
if not isinstance(bbox, list):
raise TypeError('Extent must be of type list')
# Checking that all values are either ints or floats
if not all(isinstance(n, (int, float)) for n in bbox):
raise TypeError('Bounds values must be of type int or float')
# Checking if inplace is of type bool
if not isinstance(inplace, bool):
raise TypeError('Inplace must be of type bool')
# Creating the bounds from the bbox
if len(bbox) == 6:
minx, maxx, miny, maxy = bbox[0:4]
else:
minx, maxx, miny, maxy = bbox
# Create deep copy of gdf
if not inplace:
gdf = gdf.copy(deep=True)
# Adding XY values to gdf if they are not present yet
if np.logical_not(pd.Series(['X', 'Y']).isin(gdf.columns).all()):
gdf = extract_xy(gdf)
# Clipping the GeoDataFrame
gdf = gdf[(gdf.X >= minx) & (gdf.X <= maxx) & (gdf.Y >= miny) &
(gdf.Y <= maxy)]
# Drop geometry column
gdf = gdf.drop('geometry', axis=1)
# Create new geometry column
gdf = gpd.GeoDataFrame(gdf,
geometry=gpd.points_from_xy(gdf.X, gdf.Y),
crs='EPSG:' + str(gdf.crs.to_epsg()))
# Drop Duplicates
gdf = gdf.drop_duplicates()
return gdf
def extract_xy(gdf: gpd.geodataframe.GeoDataFrame,
inplace: bool = False) -> gpd.geodataframe.GeoDataFrame:
"""
Extracting x,y coordinates from a GeoDataFrame (Points or LineStrings) and returning a GeoDataFrame with x,y coordinates as additional columns
Args:
gdf - gpd.geodataframe.GeoDataFrame created from shape file
inplace - bool - default False -> copy of the current gdf is created
Return:
gdf - gpd.geodataframe.GeoDataFrame with appended x,y columns
"""
# Input object must be a GeoDataFrame
assert isinstance(
gdf,
gpd.geodataframe.GeoDataFrame), 'Loaded object is not a GeoDataFrame'
# Store CRS of gdf
crs = gdf.crs
# Create deep copy of gdf
if not inplace:
gdf = gdf.copy(deep=True)
# Extract x,y coordinates from point shape file
if all(gdf.geom_type == "Point"):
gdf['X'] = gdf.geometry.x
gdf['Y'] = gdf.geometry.y
# Convert MultiLineString to LineString for further processing
if all(gdf.geom_type == "MultiLineString"):
gdf = gdf.explode()
# Extract x,y coordinates from line shape file
if all(gdf.geom_type == "LineString"):
gdf['points'] = [list(geometry.coords) for geometry in gdf.geometry]
df = pd.DataFrame(gdf).explode('points')
df[['X', 'Y']] = pd.DataFrame(df['points'].tolist(), index=df.index)
gdf = gpd.GeoDataFrame(df, geometry=df.geometry, crs=crs)
# Convert dip and azimuth columns to floats
if pd.Series(['dip']).isin(gdf.columns).all():
gdf['dip'] = gdf['dip'].astype(float)
if pd.Series(['azimuth']).isin(gdf.columns).all():
gdf['azimuth'] = gdf['azimuth'].astype(float)
# Convert formation column to string
if pd.Series(['formation']).isin(gdf.columns).all():
gdf['formation'] = gdf['formation'].astype(str)
return gdf
def create_linestring_gdf(
gdf: gpd.geodataframe.GeoDataFrame) -> gpd.geodataframe.GeoDataFrame:
"""
Create LineStrings from Points
Args:
gdf: GeoDataFrame containing the points of intersections between topographic contours and layer boundaries
Return:
gdf_linestring: GeoDataFrame containing LineStrings
"""
# Checking if gdf is of type GeoDataFrame
if not isinstance(gdf, gpd.geodataframe.GeoDataFrame):
raise TypeError('gdf must be of type GeoDataFrame')
# Checking geometry type of GeoDataFrame
if not all(gdf.geom_type == 'Point'):
raise ValueError(
'All objects of the GeoDataFrame must be of geom_type point')
# Checking if X and Y values are in column
if np.logical_not(pd.Series(['formation', 'Z']).isin(gdf.columns).all()):
raise ValueError('formation or Z column missing in GeoDataFrame')
# Create copy of gdf
gdf_new = gdf.copy(deep=True)
# Sort by Z values
gdf_new = gdf_new.sort_values('Z')
# Creae empty LineString list
linestrings = []
# Create LineStrings and append to list
for i in gdf_new['formation'].unique().tolist():
for j in gdf_new['Z'].unique().tolist():
linestring = create_linestring(gdf_new, i, j)
linestrings.append(linestring)
# Create gdf
gdf_linestrings = gpd.GeoDataFrame(geometry=linestrings)
# Add Z values
gdf_linestrings['Z'] = gdf_new['Z'].unique()
# Add formation name
gdf_linestrings['formation'] = gdf['formation'].unique()[0]
return gdf_linestrings
def create_linestring(
gdf: gpd.geodataframe.GeoDataFrame, formation: str,
altitude: Union[int, float]) -> shapely.geometry.linestring.LineString:
"""
Create a linestring object from a GeoDataFrame containing surface points at a given altitude and for a given
formation
Args:
gdf: GeoDataFrame containing the points of intersections between topographic contours and layer boundaries
formation: str/name of the formation
altitude: int/float value of the altitude of the points
Return:
linestring: shapely.geometry.linestring.LineString containing a LineString object
"""
# Checking if gdf is of type GeoDataFrame
if not isinstance(gdf, gpd.geodataframe.GeoDataFrame):
raise TypeError('gdf must be of type GeoDataFrame')
# Checking geometry type of GeoDataFrame
if not all(gdf.geom_type == 'Point'):
raise ValueError(
'All objects of the GeoDataFrame must be of geom_type point')
# Checking if X and Y values are in column
if np.logical_not(pd.Series(['formation', 'Z']).isin(gdf.columns).all()):
raise ValueError('formation or Z column missing in GeoDataFrame')
# Checking if the formation is of type string
if not isinstance(formation, str):
raise TypeError('formation must be of type string')
# Checking if the altitude is of type int or float
if not isinstance(altitude, (int, float)):
raise TypeError('altitude must be of type int or float')
# Creating a copy of the GeoDataFrame
gdf_new = gdf.copy(deep=True)
# Filtering GeoDataFrame by formation and altitude
gdf_new = gdf_new[gdf_new['formation'] == formation]
gdf_new = gdf_new[gdf_new['Z'] == altitude]
# Creating LineString from all available points
linestring = LineString(gdf_new.geometry.to_list())
return linestring
def load_surface_colors(path: str,
gdf: gpd.geodataframe.GeoDataFrame) -> List[str]:
"""
Load surface colors from a qml file and store the color values as list to be displayed with gpd plots
Args:
path: str/path to the qml file
gdf: GeoDataFrame of which objects are supposed to be plotted, usually loaded from a polygon/line shape file
Return:
cols: list of color values for each surface
"""
# Checking that the path is of type str
if not isinstance(path, str):
raise TypeError('path must be provided as string')
# Checking that the gdf is of type GeoDataFrame
if not isinstance(gdf, gpd.geodataframe.GeoDataFrame):
raise TypeError('object must be of type GeoDataFrame')
# Parse qml
column, classes = parse_categorized_qml(path)
# Create style dict
style_df = pd.DataFrame(build_style_dict(classes)).transpose()
# Create deep copy of gdf
gdf_copy = gdf.copy(deep=True)
# Append style_df to copied gdf
gdf_copy["Color"] = gdf_copy[column].replace(style_df.color.to_dict())
# Sort values of gdf by provided column, usually the formation
gdf_copy = gdf_copy.sort_values(column)
# Filter for unique formations
gdf_copy = gdf_copy.groupby([column], as_index=False).last()
# Create list of remaining colors
cols = gdf_copy['Color'].to_list()
return cols
def clip_by_shape(gdf: gpd.geodataframe.GeoDataFrame,
shape: gpd.geodataframe.GeoDataFrame,
inplace: bool = False) -> gpd.geodataframe.GeoDataFrame:
"""
Clipping vector data by extent
Args:
gdf: GeoDataFrame to be clipped
shape: GeoDataFrame acting as bbox
inplace: - bool - default False -> copy of the current gdf is created
Return:
gdf: GeoDataFrame with the clipped values
"""
# Checking if the gdf is of type GeoDataFrame
if not isinstance(gdf, gpd.geodataframe.GeoDataFrame):
raise TypeError('gdf must be of type GeoDataFrame')
# Checking if the shape is of type GeoDataFrame
if not isinstance(shape, gpd.geodataframe.GeoDataFrame):
raise TypeError('shape must be of type GeoDataFrame')
# Checking if inplace is of type bool
if not isinstance(inplace, bool):
raise TypeError('Inplace must be of type bool')
# Create deep copy of gdf
if not inplace:
gdf = gdf.copy(deep=True)
# Setting the extent
extent = set_extent(gdf=shape)
# Clipping the gdf
gdf = clip_by_extent(gdf, extent, inplace=inplace)
return gdf
def extract_z(gdf: gpd.geodataframe.GeoDataFrame,
dem: Union[np.ndarray, rasterio.io.DatasetReader],
inplace: bool = False,
**kwargs) -> gpd.geodataframe.GeoDataFrame:
"""
Extracting altitude values from digital elevation model
Args:
gdf - gpd.geodataframe.GeoDataFrame containing x,y values
dem - rasterio.io.DatasetReader containing the z values
inplace - bool - default False -> copy of the current gdf is created
Kwargs:
extent - list containing the extent of the np.ndarray, must be provided in the same CRS as the gdf
Return:
gdf - gpd.geodataframe.GeoDataFrame containing x,y,z values obtained from a DEM
"""
# Input object must be a GeoDataFrame
if not isinstance(gdf, gpd.geodataframe.GeoDataFrame):
raise TypeError('Loaded object is not a GeoDataFrame')
# Create deep copy of gdf
if not inplace:
gdf = gdf.copy(deep=True)
# Input object must be a np.ndarray or a rasterio.io.DatasetReader
if not isinstance(dem, (np.ndarray, rasterio.io.DatasetReader)):
raise TypeError(
'Loaded object is not a np.ndarray or rasterio.io.DatasetReader')
# The GeoDataFrame must not contain a Z-column
if pd.Series(['Z']).isin(gdf.columns).all():
raise ValueError('Data already contains Z-values')
# Extracting z values from a DEM loaded with Rasterio
if isinstance(dem, rasterio.io.DatasetReader):
try:
if gdf.crs == dem.crs:
if np.logical_not(
pd.Series(['X', 'Y']).isin(gdf.columns).all()):
gdf = extract_xy(gdf)
gdf['Z'] = [
z[0] for z in dem.sample(gdf[['X', 'Y']].to_numpy())
]
else:
crs_old = gdf.crs
gdf = gdf.to_crs(crs=dem.crs)
gdf = extract_xy(gdf)
gdf['Z'] = [
z[0] for z in dem.sample(gdf[['X', 'Y']].to_numpy())
]
gdf = gdf.to_crs(crs=crs_old)
del gdf['X']
del gdf['Y']
gdf = extract_xy(gdf)
except IndexError:
raise ValueError(
'One or more points are located outside the boundaries of the raster'
)
# Extracting z values from a DEM as np.ndarray
else:
if np.logical_not(pd.Series(['X', 'Y']).isin(gdf.columns).all()):
gdf = extract_xy(gdf)
extent = kwargs.get('extent', None)
assert extent is not None, 'Extent of array is needed to extract Z values'
gdf['Z'] = [
sample(dem, extent, gdf[['X', 'Y']].values.tolist()[i])
for i, point in enumerate(gdf[['X', 'Y']].values.tolist())
]
# Convert dip and azimuth columns to floats
if pd.Series(['dip']).isin(gdf.columns).all():
gdf['dip'] = gdf['dip'].astype(float)
if pd.Series(['azimuth']).isin(gdf.columns).all():
gdf['azimuth'] = gdf['azimuth'].astype(float)
# Convert formation column to string
if pd.Series(['formation']).isin(gdf.columns).all():
gdf['formation'] = gdf['formation'].astype(str)
return gdf
def extract_coordinates(gdf: gpd.geodataframe.GeoDataFrame,
dem: Union[np.ndarray, rasterio.io.DatasetReader,
type(None)] = None,
inplace: bool = False,
**kwargs) -> gpd.geodataframe.GeoDataFrame:
"""
Extract x,y and z coordinates from a GeoDataFrame
Args:
gdf - gpd.geodataframe.GeoDataFrame containing Points or LineStrings
dem - rasterio.io.DatasetReader containing the z values
Kwargs:
extent - list containing the extent of the np.ndarray, must be provided in the same CRS as the gdf
Return:
gdf - gpd.geodataframe.GeoDataFrame containing x, y and z values
"""
# Input object must be a GeoDataFrame
if not isinstance(gdf, gpd.geodataframe.GeoDataFrame):
raise TypeError('Loaded object is not a GeoDataFrame')
# Create deep copy of gdf
if not inplace:
gdf = gdf.copy(deep=True)
# Checking if Z is in GeoDataFrame
if np.logical_not(pd.Series(['Z']).isin(gdf.columns).all()):
# Checking if dem is not None
if dem is None:
raise ValueError('DEM is missing')
# Checking if DEM is of type np.ndarray or rasterio object
if not isinstance(dem, (np.ndarray, rasterio.io.DatasetReader)):
raise TypeError(
'Loaded object is not a np.ndarray or Rasterio object')
extent = kwargs.get('extent', None)
# Checking if X and Y column already exist in gdf
if np.logical_not(pd.Series(['X', 'Y']).isin(gdf.columns).all()):
if isinstance(dem, np.ndarray):
gdf = extract_z(gdf, dem, extent=extent)
# Extract XYZ values if dem is rasterio object
else:
# Extract XYZ values if CRSs are matching
if gdf.crs == dem.crs:
gdf = extract_z(gdf, dem)
# Convert gdf before XYZ values extraction
else:
crs_old = gdf.crs
gdf = gdf.to_crs(crs=dem.crs)
gdf.rename(columns={'X': 'X1', 'Y': 'Y1'})
gdf = extract_z(extract_xy(gdf), dem)
gdf = gdf.to_crs(crs=crs_old)
del gdf['X']
del gdf['Y']
gdf.rename(columns={'X1': 'X', 'Y1': 'Y'})
else:
# Extract XYZ values if dem is of type np.ndarray
if isinstance(dem, np.ndarray):
gdf = extract_z(extract_xy(gdf), dem, extent=extent)
# Extract XYZ values if dem is rasterio object
else:
# Extract XYZ values if CRSs are matching
if gdf.crs == dem.crs:
gdf = extract_z(extract_xy(gdf), dem)
# Convert gdf before XYZ values extraction
else:
crs_old = gdf.crs
gdf = gdf.to_crs(crs=dem.crs)
gdf = extract_z(extract_xy(gdf), dem)
gdf = gdf.to_crs(crs=crs_old)
del gdf['X']
del gdf['Y']
gdf = extract_xy(gdf)
else:
# Checking if X and Y column already exist in gdf
if np.logical_not(pd.Series(['X', 'Y']).isin(gdf.columns).all()):
gdf = extract_xy(gdf, inplace=inplace)
# Convert dip and azimuth columns to floats
if pd.Series(['dip']).isin(gdf.columns).all():
gdf['dip'] = gdf['dip'].astype(float)
if pd.Series(['azimuth']).isin(gdf.columns).all():
gdf['azimuth'] = gdf['azimuth'].astype(float)
# Convert formation column to string
if pd.Series(['formation']).isin(gdf.columns).all():
gdf['formation'] = gdf['formation'].astype(str)
return gdf
