Fix breaking changes with polars 0.19.6

CHANGELOG.rst

@@ -5,6 +5,7 @@ Development
 ***********
 
 - Fix multiple issues with DwdObservationRequest API
+- Raise minimum version of polars to 0.19.6 due to breaking changes
 
 0.61.0 (06.10.2023)
 *******************

docs/img/readme_img.py

@@ -23,9 +23,9 @@ def create_temperature_ts_plot():
         parameter="temperature_air_mean_200", resolution="daily", period="historical"
     ).filter_by_name("Hohenpeissenberg")
     df = stations.values.all().df
-    df_annual = df.groupby(pl.col("date").dt.year(), maintain_order=True).agg(pl.col("value").mean().alias("value"))
+    df_annual = df.group_by(pl.col("date").dt.year(), maintain_order=True).agg(pl.col("value").mean().alias("value"))
     df_annual = df_annual.with_columns(
-        pl.col("date").cast(str).str.strptime(datatype=pl.Datetime, fmt="%Y"), pl.col("value").mean().alias("mean")
+        pl.col("date").cast(str).str.to_datetime("%Y"), pl.col("value").mean().alias("mean")
     )
     fig, ax = plt.subplots(tight_layout=True)
     df.to_pandas().plot("date", "value", ax=ax, color="blue", label="Tmean,daily", legend=False)

example/observations_station_gaussian_model.py

@@ -85,7 +85,7 @@ def __init__(self, station_data: StationsResult):
 
     def get_valid_data(self, result_values: pl.DataFrame) -> pl.DataFrame:
         valid_data_lst = []
-        for _, group in result_values.groupby(pl.col("date").dt.year()):
+        for _, group in result_values.group_by(pl.col("date").dt.year()):
             if self.validate_yearly_data(group):
                 valid_data_lst.append(group)
 
@@ -110,7 +110,7 @@ def make_composite_yearly_model(self, valid_data: pl.DataFrame) -> Tuple[Gaussia
         index_per_year = x.max() / number_of_years
 
         pars, composite_model = None, None
-        for year, group in valid_data.groupby(pl.col("date").dt.year(), maintain_order=True):
+        for year, group in valid_data.group_by(pl.col("date").dt.year(), maintain_order=True):
             gmod = GaussianModel(prefix=f"g{year}_")
             if pars is None:
                 pars = gmod.make_params()

example/wetterdienst_notebook.ipynb

@@ -760,7 +760,7 @@
     "        agg_df = group.groupby(pl.col(\"date\").dt.year(), maintain_order=True).agg([pl.col(\"value\").mean()])\n",
     "    agg_df = agg_df.select(\n",
     "        pl.lit(parameter).alias(\"parameter\"),\n",
-    "        pl.col(\"date\").cast(pl.Utf8).str.strptime(datatype=pl.Datetime, fmt=\"%Y\"),\n",
+    "        pl.col(\"date\").cast(pl.Utf8).str.to_datetime(\"%Y\"),\n",
     "        pl.col(\"value\"),\n",
     "    )\n",
     "    data.append(agg_df)\n",

pyproject.toml

@@ -119,7 +119,7 @@ numpy = ">=1.22,<1.27"
 pandas = ">=2,<2.2"
 Pint = ">=0.17,<0.23"
 platformdirs = "<4"
-polars = ">=0.16,<0.20"
+polars = ">=0.19.6,<0.20"
 pyarrow = ">=13,<14"
 pypdf = ">=3.12.1,<3.17"
 python-dateutil = "<3"

tests/core/timeseries/test_io.py

@@ -613,7 +613,7 @@ def test_export_duckdb(settings_si_false, tmp_path):
         settings=settings_si_false,
     ).filter_by_station_id(station_id=[1048])
     filename = tmp_path.joinpath("test.duckdb")
-    df = request.values.all().df
+    df = request.values.all().df.sort(["parameter", "date"])
     ExportMixin(df=df).to_target(f"duckdb:///{filename}?table=testdrive")
     connection = duckdb.connect(str(filename), read_only=True)
     cursor = connection.cursor()
@@ -624,9 +624,9 @@ def test_export_duckdb(settings_si_false, tmp_path):
     assert results[300_000] == (
         "01048",
         "climate_summary",
-        "temperature_air_max_200",
+        "temperature_air_min_200",
         dt.datetime(1939, 7, 26),
-        21.0,
+        10.0,
         1.0,
     )