cleaning some code duplication (#256)

* cleaning somd code duplication

* update whats new

Author: larsbuntemeyer

docs/whats_new.rst

@@ -11,7 +11,15 @@ What's New
 v0.8.3 (22 October 2025)
 ------------------------
 
-Bugfix release.
+Maintenance and bugfix release.
+
+Internal Changes
+~~~~~~~~~~~~~~~~
+
+- Cleaning of unused imports (:pull:`256`)
+
+Bugfixes
+~~~~~~~~
 
 - Fix fetching of vertical coordinates from CF models (:pull:`255`)
 

pyremo/preproc/__init__.py

@@ -9,7 +9,8 @@
     CFPreprocessor,
     RemoPreprocessor,
 )
-from .remapping import remap, remap_remo, to_netcdf, to_tar
+from .remapping import remap, remap_remo
+from .utils import to_netcdf, to_tar
 
 # from .remap_new import Remapper
 from .utils import write_forcing_file

pyremo/preproc/core.py

@@ -222,56 +222,56 @@ def open_datasets(datasets, ref_ds=None, time_range=None):
     return output
 
 
-def gfile(ds, ref_ds=None, tos=None, time_range=None, attrs=None):
-    """Creates a global dataset ready for preprocessing.
-
-    This function creates a homogenized global dataset. If neccessary,
-    units are converted and the sea surface temperature ``tos`` is
-    interpolated spatially and temporally to the atmospheric grid.
-
-    Parameters
-    ----------
-    ds : xarray.Dataset or dict of filenames
-        Input dataset from a global model according to CF conventions.
-
-    ref_ds : xarray.Dataset
-        Reference datasets that is used for determining the grid and vertical
-        coordinates and the global attributes. If ``ref_ds=None``, ``ta`` from
-        the input dataset is used as a reference.
-
-    tos : xarray.Dataset
-        Sea surface dataset.
-
-    time_range :
-        The common time range from the input and sst that should be used.
-
-    attrs:
-        Global attributes for the output dataset. If ``attrs=None``, the global
-        attributes from ``ref_ds`` are used.
-
-    Returns
-    -------
-    gfile : xarray.Dataset
-        Global dataset ready for preprocessing.
-
-    """
-    if isinstance(ds, dict):
-        ds = open_datasets(ds, ref_ds, time_range)
-    else:
-        ds = ds.copy()
-        ds["akgm"], ds["bkgm"] = get_vc(ds)
-        ds = check_lev(ds)
-
-    if time_range is None:
-        time_range = ds.time
-        ds = ds.sel(time=time_range)
-
-    if tos is not None:
-        ds["tos"] = map_sst(tos, ds.sel(time=time_range))
-    # ds = ds.rename({"lev": lev_gm})
-    ds = convert_units(ds)
-    # if "sftlf" in ds:
-    #    ds["sftlf"] = np.around(ds.sftlf)
-    if attrs is None:
-        attrs = ds.attrs
-    return ds
+# def gfile(ds, ref_ds=None, tos=None, time_range=None, attrs=None):
+#     """Creates a global dataset ready for preprocessing.
+
+#     This function creates a homogenized global dataset. If neccessary,
+#     units are converted and the sea surface temperature ``tos`` is
+#     interpolated spatially and temporally to the atmospheric grid.
+
+#     Parameters
+#     ----------
+#     ds : xarray.Dataset or dict of filenames
+#         Input dataset from a global model according to CF conventions.
+
+#     ref_ds : xarray.Dataset
+#         Reference datasets that is used for determining the grid and vertical
+#         coordinates and the global attributes. If ``ref_ds=None``, ``ta`` from
+#         the input dataset is used as a reference.
+
+#     tos : xarray.Dataset
+#         Sea surface dataset.
+
+#     time_range :
+#         The common time range from the input and sst that should be used.
+
+#     attrs:
+#         Global attributes for the output dataset. If ``attrs=None``, the global
+#         attributes from ``ref_ds`` are used.
+
+#     Returns
+#     -------
+#     gfile : xarray.Dataset
+#         Global dataset ready for preprocessing.
+
+#     """
+#     if isinstance(ds, dict):
+#         ds = open_datasets(ds, ref_ds, time_range)
+#     else:
+#         ds = ds.copy()
+#         ds["akgm"], ds["bkgm"] = get_vc(ds)
+#         ds = check_lev(ds)
+
+#     if time_range is None:
+#         time_range = ds.time
+#         ds = ds.sel(time=time_range)
+
+#     if tos is not None:
+#         ds["tos"] = map_sst(tos, ds.sel(time=time_range))
+#     # ds = ds.rename({"lev": lev_gm})
+#     ds = convert_units(ds)
+#     # if "sftlf" in ds:
+#     #    ds["sftlf"] = np.around(ds.sftlf)
+#     if attrs is None:
+#         attrs = ds.attrs
+#     return ds

pyremo/preproc/remapping.py

@@ -1,5 +1,3 @@
-import os
-
 import cf_xarray as cfxr  # noqa
 import cordex as cx
 import numpy as np
@@ -8,7 +6,7 @@
 import pyremo as pr
 
 from . import physics
-from .constants import fillvars, lev_input
+from .constants import lev_input
 from .core import const, get_akbkem
 from .utils import update_attrs
 from .xpyintorg import (
@@ -40,65 +38,6 @@
 xr.set_options(keep_attrs=True)
 
 
-def get_filename(date, expid="000000", template=None):
-    if template is None:
-        template = "x{}x{}.nc"
-    return template.format(expid, date.strftime(format="%Y%m%d%H"))
-
-
-def to_netcdf(
-    ads,
-    path="",
-    expid="000000",
-    template=None,
-    tempfiles=None,
-    missval=1.0e20,
-    **kwargs,
-):
-    """write dataset to netcdf
-
-    by default, each timestep goes into a separate output file
-
-    """
-    if not os.path.isdir(path):
-        os.makedirs(path)
-    expand_time = [var for var, da in ads.items() if "time" in da.dims]
-    if template is None:
-        template = "a{}a{}.nc"
-    dates, datasets = zip(*ads.groupby("time"))
-    paths = [os.path.join(path, get_filename(date, expid, template)) for date in dates]
-    dsets = []
-    # expand time dimension only for variables not coordinates.
-    for ds in datasets:
-        for var, da in ds.items():
-            if var in expand_time:
-                ds[var] = da.expand_dims("time")
-            if var in fillvars:
-                ds[var].encoding["_FillValue"] = missval
-            else:
-                ds[var].encoding["_FillValue"] = None
-        dsets.append(ds)
-    # dsets = [dset.expand_dims('time') for dset in datasets]
-    return xr.save_mfdataset(dsets, paths, **kwargs)
-
-
-def to_tar(files, tar_file, mode="w"):
-    import tarfile
-
-    try:
-        from tqdm import tqdm
-    except Exception:
-
-        def tqdm(x):
-            return x
-
-    tf = tarfile.open(tar_file, mode=mode)
-    for f in tqdm(files, desc="creating tarfile"):
-        tf.add(f, arcname=os.path.basename(f), recursive=False)
-    tf.close()
-    return tar_file
-
-
 def broadcast_coords(ds, coords=("lon", "lat")):
     """broadcast 1d global coordinates"""
     lat2d, lon2d = xr.broadcast(ds[coords[1]], ds[coords[0]])

pyremo/preproc/utils.py

@@ -207,7 +207,6 @@ def to_netcdf(
     path="",
     expid="000000",
     template=None,
-    tempfiles=None,
     missval=1.0e20,
     **kwargs,
 ):
@@ -242,11 +241,10 @@ def to_netcdf(
         DeprecationWarning,
         stacklevel=2,
     )
-    if not os.path.isdir(path):
-        os.makedirs(path)
-    expand_time = [var for var, da in ads.items() if "time" in da.dims]
-    if template is None:
-        template = "a{}a{}.nc"
+    os.makedirs(path, exist_ok=True)
+    expand_time = [
+        var for var, da in ads.items() if "time" in da.dims and "time" not in da.coords
+    ]
     dates, datasets = zip(*ads.groupby("time"))
     paths = [os.path.join(path, get_filename(date, expid, template)) for date in dates]
     dsets = []
@@ -262,9 +260,6 @@ def to_netcdf(
         dsets.append(ds)
     # dsets = [dset.expand_dims('time') for dset in datasets]
     xr.save_mfdataset(dsets, paths, **kwargs)
-    if tempfiles is not None:
-        for f in tempfiles:
-            os.remove(f)
     return paths