diff --git a/src/toargridding/grids.py b/src/toargridding/grids.py
index 745599dfb9395b60de1a85d3261cd4d243fa54d5..afc47ed16d1b371f3b629506eb52840642991854 100644
--- a/src/toargridding/grids.py
+++ b/src/toargridding/grids.py
@@ -63,8 +63,8 @@ class GridDefinition(ABC):
"""
@abstractmethod
- def as_xarray(timeseries: dict[str, pd.DataFrame], metadata: pd.DataFrame) -> dict[str, xr.Dataset]:
- """conversion of panda Dataframes to an xarray dataset
+ def as_xarray(self, data: AnalysisRequestResult) -> xr.Dataset:
+ """gridding of a request to the TOAR database
This includes the required setup to store the results as netCDF file according to CF (https://cfconventions.org/cf-conventions/cf-conventions.html)
"""
@@ -105,8 +105,12 @@ class RegularGrid(GridDefinition):
super().__init__()
# TODO make sure only sensible resolutions
- self.lat = Coordinate.from_resolution(Coordinates.latitude, lat_resolution, min=-90, max=90, wraps=False)
- self.lon = Coordinate.from_resolution(Coordinates.longitude, lon_resolution, min=-180, max=180, wraps=True)
+ self.lat = Coordinate.from_resolution(
+ Coordinates.latitude, lat_resolution, min=-90, max=90, wraps=False
+ )
+ self.lon = Coordinate.from_resolution(
+ Coordinates.longitude, lon_resolution, min=-180, max=180, wraps=True
+ )
spatial_shape = (self.lon.size, self.lat.size)
spatial_size = self.lon.size * self.lat.size
self.dims = [
@@ -115,7 +119,9 @@ class RegularGrid(GridDefinition):
Coordinates.longitude.name,
]
- self._as_xy_index = np.dstack(np.meshgrid(range(self.lat.size), range(self.lon.size))).reshape(-1, 2)
+ self._as_xy_index = np.dstack(
+ np.meshgrid(range(self.lat.size), range(self.lon.size))
+ ).reshape(-1, 2)
self._as_i_index = np.arange(spatial_size).reshape(spatial_shape).T
@property
@@ -135,13 +141,17 @@ class RegularGrid(GridDefinition):
results of the request, including data, station coordinates and metadata of request
"""
- data_grouped_by_cell = self.group_data_by_cell(data.stations_data, data.stations_coords)
+ data_grouped_by_cell = self.group_data_by_cell(
+ data.stations_data, data.stations_coords
+ )
cell_statistics = self.get_cell_statistics(data_grouped_by_cell)
dataset = self.create_dataset(cell_statistics, data.metadata)
return dataset
- def group_data_by_cell(self, data: pd.DataFrame, coords: pd.DataFrame) -> DataFrameGroupBy:
+ def group_data_by_cell(
+ self, data: pd.DataFrame, coords: pd.DataFrame
+ ) -> DataFrameGroupBy:
"""grouping of stations into cells
This function converts the lat/lon coordinates of the stations into cell indices and groups stations belonging to one cell.
@@ -157,7 +167,9 @@ class RegularGrid(GridDefinition):
cell_indices = self.as_cell_index(coords)
# will convert cell_indices to float as some nans ar present
- data_with_indices = data.join(cell_indices.to_frame(GridDefinition.cell_index_name), how="outer")
+ data_with_indices = data.join(
+ cell_indices.to_frame(GridDefinition.cell_index_name), how="outer"
+ )
return data_with_indices.groupby(GridDefinition.cell_index_name)
@@ -180,7 +192,9 @@ class RegularGrid(GridDefinition):
return stats
- def create_dataset(self, cell_statistics: dict[str, pd.DataFrame], metadata: Metadata) -> xr.Dataset:
+ def create_dataset(
+ self, cell_statistics: dict[str, pd.DataFrame], metadata: Metadata
+ ) -> xr.Dataset:
"""creation of data set and filling with results from the gridding
Parameters:
@@ -202,7 +216,9 @@ class RegularGrid(GridDefinition):
gridded_ds = self.get_empty_grid(time, metadata)
for variable, aggregated_data in cell_statistics.items():
- data_array_dict = self.get_data_array_dict(time, aggregated_data, variable, metadata)
+ data_array_dict = self.get_data_array_dict(
+ time, aggregated_data, variable, metadata
+ )
gridded_ds = gridded_ds.assign(data_array_dict)
return gridded_ds
@@ -232,7 +248,9 @@ class RegularGrid(GridDefinition):
gridded_variable = Variable.from_data(gridded_data, variable, metadata)
return {variable.name: gridded_variable.as_data_array(self.dims)}
- def create_gridded_data(self, time: Coordinate, grouped_timeseries: pd.DataFrame) -> np.array:
+ def create_gridded_data(
+ self, time: Coordinate, grouped_timeseries: pd.DataFrame
+ ) -> np.array:
"""converts the available cell data to a full lat/lon-temporal data cube.
Parameters:
@@ -250,7 +268,9 @@ class RegularGrid(GridDefinition):
values[...] = self.fill_value
index = self._as_xy_index[grouped_timeseries.index.astype(int)]
- values[:, index.T[0], index.T[1]] = grouped_timeseries.values.reshape(-1, time.size).T
+ values[:, index.T[0], index.T[1]] = grouped_timeseries.values.reshape(
+ -1, time.size
+ ).T
return values
@@ -258,13 +278,17 @@ class RegularGrid(GridDefinition):
"""converts coordinates of stations into x and y indices of the regular grid"""
id_x = self.coord_to_index(coords[self.lat.name], self.lat.min, self.lat.step)
- id_y = self.coord_to_index(coords[self.lon.name], self.lon.min, self.lon.step, len(self.lon.data))
+ id_y = self.coord_to_index(
+ coords[self.lon.name], self.lon.min, self.lon.step, len(self.lon.data)
+ )
id_i = self._as_i_index[id_x, id_y]
return pd.Series(id_i, index=id_x.index)
- def coord_to_index(self, coord: pd.Series, x0_axis: float, d_axis: float, maxBin4Wrap: int = None) -> np.array:
+ def coord_to_index(
+ self, coord: pd.Series, x0_axis: float, d_axis: float, maxBin4Wrap: int = None
+ ) -> np.array:
"""converts a coordinate into a bin index on one axis
Parameters:
@@ -284,7 +308,9 @@ class RegularGrid(GridDefinition):
ids[ids < 0] += maxBin4Wrap
return ids
- def get_empty_grid(self, time: Variable, metadata: Metadata) -> xr.Dataset: # TODO make CF-compliant => docs
+ def get_empty_grid(
+ self, time: Variable, metadata: Metadata
+ ) -> xr.Dataset: # TODO make CF-compliant => docs
"""creation of an empty dataset without data
Sets up a dataset with its three axis: time, longitude and latitude.