add data loaders

climate_toolbox/io/io.py

@@ -3,69 +3,113 @@
 from climate_toolbox.utils.utils import *
 
 
-def standardize_climate_data(ds):
-    """
-    Read climate data and standardize units to:
-        - lon and lat,
-        - lon to -180 to 180 and
+def load_climate_data(data_type, file_path):
+    """ load_climate_data(data_type, file_path)
+    Read and prepare climate data
+
+    :param data_type: str
+        datatype to be read, supported types:
+        bcsd, gmfd, best, era5
 
-    Parameters
-    ----------
-    ds:  xr.Dataset
+    :param file_path: str
+        File path
 
-    Returns
-    -------
-    xr.Dataset
+    :return: ds: xr.Dataset
+         xarray dataset loaded in memory
     """
 
-    ds = rename_coords_to_lon_and_lat(ds)
-    ds = convert_lons_split(ds, lon_name='lon')
+    return _load_climate_data(
+        _find_loader(data_type),
+        file_path
+    )
+
+
+def _load_climate_data(loader, file_path):
+    with xr.open_dataset(file_path) as ds:
+        ds.load()
+
+    return loader(ds)
 
-    return ds
 
+def load_min_max_temperatures(data_type, file_path_tmin, file_path_tmax):
+    """ load_min_max_temperatures(data_type, file_path_tmin, file_path_tmax)
 
-def load_bcsd(fp, varname, lon_name='lon', broadcast_dims=('time',)):
+    :param data_type: str
+        datatype to be read, supported types:
+        bcsd, gmfd, best, era5
+
+    :param file_path_tmin: path for min temperature
+    :param file_path_tmax: path for max temperature
+    :return:
+        ds_tasmax: xr.Dataset, ds_tasmin: xr.Dataset
     """
-    Read and prepare climate data
 
-    After reading data, this method also fills NA values using linear
-    interpolation, and standardizes longitude to -180:180
+    return _load_min_max_temperatures(
+        _find_loader(data_type),
+        file_path_tmin,
+        file_path_tmax
+    )
 
-    Parameters
-    ----------
-    fp: str
-        File path or dataset
 
-    varname: str
-        Variable name to be read
+def _load_min_max_temperatures(loader, file_path_tmin, file_path_tmax):
+    with xr.open_dataset(file_path_tmin) as ds_tasmin:
+        ds_tasmin.load()
+    with xr.open_dataset(file_path_tmax) as ds_tasmax:
+        ds_tasmax.load()
 
-    lon_name : str, optional
-        Name of the longitude dimension (defualt selects from ['lon' or
-        'longitude'])
+    return loader(ds_tasmin), loader(ds_tasmax)
 
-    Returns
-    -------
-    xr.Dataset
-         xarray dataset loaded into memory
-    """
 
-    if lon_name is not None:
-        lon_names = [lon_name]
+def _find_loader(data_type):
+    """ Helper function to find climate data loader """
 
-    if hasattr(fp, 'sel_points'):
-        ds = fp
+    data_type = data_type.lower()
 
+    if 'bcsd' in data_type:
+        loader = load_bcsd
+    elif 'gmfd' in data_type:
+        loader = load_gmfd
+    elif 'best' in data_type:
+        loader = load_best
+    elif 'era' in data_type:
+        loader = load_era5
     else:
-        with xr.open_dataset(fp) as ds:
-            ds.load()
+        raise TypeError("'" + data_type + "' not supported. Supported data "
+                                          "types are: NASA BCSD, GMFD, BEST, ERA5.")
+    return loader
+
 
-    return standardize_climate_data(ds)
+def standardize_climate_data(ds):
+    """ standardize_climate_data(ds)
+    Standardize climate data units to:
+        - lon and lat,
+        - lon to -180 to 180 and
+
+    :param ds: xr.Dataset
+    :return: ds: xr.Dataset
+    """
 
+    ds = rename_coords_to_lon_and_lat(ds)
+    ds = convert_lons_split(ds, lon_name='lon')
 
-def load_gmfd(fp, varname, lon_name='lon', broadcast_dims=('time',)):
-    pass
+    return ds
+
+
+def load_bcsd(ds):
+    return ds
 
 
-def load_best(fp, varname, lon_name='lon', broadcast_dims=('time',)):
+def load_gmfd(ds):
+    if 'tmin' in ds.data_vars or 'tmax' in ds.data_vars:
+        return standardize_climate_data(ds)
+    if 'lat' not in ds.coords or 'lon' not in ds.coords:
+        ds = rename_coords_to_lon_and_lat(ds)
+    return convert_lons_split(ds, lon_name="lon")
+
+
+def load_best():
     pass
 
+
+def load_era5():
+    pass