gridmarthe.operasem.gis

GIS utility for marthe grids

gridmarthe.operasem.gis.to_geodataframe(ds, epsg='EPSG:27572', fmt='long')[source]

Convert marthegrid.Dataset to a geodataframe

Parameters:

  • ds (xarray.Dataset) – The dataset to convert.

  • epsg (str, optional) – The EPSG code for the coordinate reference system, by default ‘EPSG:27572’.

  • fmt (str, optional) – The format of the output GeoDataFrame, either ‘long’ or ‘wide’, by default ‘long’.

Returns:

geopandas.GeoDataFrame – The converted GeoDataFrame.

gridmarthe.operasem.gis.clip_dataset(ds, gdf, crs=27572, engine='gdf')[source]

Clip a xarray Dataset with a gpd.GeoDataFrame

Needs rioxarray. If not installed, raise ModuleNotFoundError please install it or reinstall gridmarthe with optional dependancies: pip install gridmarthe[opt]

See: https://corteva.github.io/rioxarray/html/examples/clip_geom.html Todo: shapely version

Parameters:

  • ds (xarray.Dataset) – The dataset to clip.

  • gdf (geopandas.GeoDataFrame) – The GeoDataFrame containing the geometry to clip with.

  • crs (int, optional) – The coordinate reference system to use for the clipping, by default 27572.

  • engine (str, optional) – The engine to use for clipping, by default ‘gdf’.

Returns:

xarray.DataArray or xarray.Dataset – The clipped dataset.

gridmarthe.operasem.gis.subset_with_coords(da, dims=['x', 'y'], gdf=None, xmin=None, ymin=None, xmax=None, ymax=None)[source]

subset DataArray or Dataset on rectangular shape, with gpd.GeoDataFrame or bounds

Parameters:

  • da (xarray.DataArray or xarray.Dataset) – The data to subset.

  • dims (list of str, optional) – The dimensions to use for subsetting, by default [‘x’, ‘y’].

  • gdf (geopandas.GeoDataFrame, optional) – A GeoDataFrame containing the geometry to use for subsetting, by default None.

  • xmin (float, optional) – The manual bounds to use for subsetting, by default None.

  • ymin (float, optional) – The manual bounds to use for subsetting, by default None.

  • xmax (float, optional) – The manual bounds to use for subsetting, by default None.

  • ymax (float, optional) – The manual bounds to use for subsetting, by default None.

Returns:

xarray.DataArray or xarray.Dataset – The subsetted data.

gridmarthe.operasem.gis.transf_proj(ds, from_epsg='EPSG:27572', to_epsg='EPSG:2154', engine='rioxarray')[source]

Transform grid of a dataset using pyproj

engine: rioxarray, meshgrid, xy

Data (ds) needs to be a 2D array with xy as coords dimensions and time sliced. If needed, use gridmarthe.assign_coords() first.

Warning: IN DEVELOPMENT, not tested yet // USE WITH CAUTION

gridmarthe.operasem.gis.to_raster(ds, varname=None, x_dim='x', y_dim='y', time=None, epsg=27572, filename_tpl='raster')[source]

Write a xr.DataArray to a raster file

Notes

  • Warning, only functionnal for regular grids

  • Requires rasterio/rioxarray packages

  • TODO: add support for irregular grids, using PostMARTHE QGIS plugin code.

Parameters:

  • ds (xarray.Dataset, xarray.DataArray) – The dataset/dataArray to write to a raster file

  • x_dim (str, optional) – The name of the x dimension, by default ‘x’.

  • y_dim (str, optional) – The name of the y dimension, by default ‘y’.

  • time (str, list) – time or list of time from da.time

  • epsg (int, optional) – The EPSG code for the coordinate reference system, by default 27572.

  • filename_tpl (str, optional) – The output file template for the raster file, by default ‘raster’. Final name will be ‘{filename_tpl}_{time}.tiff’

Returns:

None – The function writes the raster file and returns None.