add order for polynomial interpolation, fixes #8762 (#9079)

* add order for polynomial, fixes #8762

* add bugfixes in whats_new.rst, fixes #8762

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* fix typo

* Update xarray/core/resample.py

setdefault "bounds_error", and not forcing it. Thanks @dcherian

Co-authored-by: Deepak Cherian <dcherian@users.noreply.github.com>

* chore(test) : polynomial resample

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* fix(setdefault) : change dict to arg

* fix(test) : polynomial interpolate tests

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* chore(test) : add comment using interp1d for polynomial

* Update xarray/tests/test_groupby.py

* Update doc/whats-new.rst

* Update whats-new.rst

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Deepak Cherian <dcherian@users.noreply.github.com>

doc/whats-new.rst

@@ -41,6 +41,10 @@ Deprecations
 Bug fixes
 ~~~~~~~~~
 
+- :py:meth:`DataArrayResample.interpolate` and :py:meth:`DatasetResample.interpolate` method now
+  support aribtrary kwargs such as ``order`` for polynomial interpolation. (:issue:`8762`).
+  By `Nicolas Karasiak <https://github.com/nkarasiak>`_.
+
 
 Documentation
 ~~~~~~~~~~~~~

xarray/core/resample.py

@@ -140,7 +140,7 @@ def nearest(self, tolerance: float | Iterable[float] | None = None) -> T_Xarray:
             {self._dim: grouper.full_index}, method="nearest", tolerance=tolerance
         )
 
-    def interpolate(self, kind: InterpOptions = "linear") -> T_Xarray:
+    def interpolate(self, kind: InterpOptions = "linear", **kwargs) -> T_Xarray:
         """Interpolate up-sampled data using the original data as knots.
 
         Parameters
@@ -168,17 +168,18 @@ def interpolate(self, kind: InterpOptions = "linear") -> T_Xarray:
         scipy.interpolate.interp1d
 
         """
-        return self._interpolate(kind=kind)
+        return self._interpolate(kind=kind, **kwargs)
 
-    def _interpolate(self, kind="linear") -> T_Xarray:
+    def _interpolate(self, kind="linear", **kwargs) -> T_Xarray:
         """Apply scipy.interpolate.interp1d along resampling dimension."""
         obj = self._drop_coords()
         (grouper,) = self.groupers
+        kwargs.setdefault("bounds_error", False)
         return obj.interp(
             coords={self._dim: grouper.full_index},
             assume_sorted=True,
             method=kind,
-            kwargs={"bounds_error": False},
+            kwargs=kwargs,
         )
 
 

xarray/tests/test_groupby.py

@@ -2074,13 +2074,18 @@ def test_upsample_interpolate(self) -> None:
             "slinear",
             "quadratic",
             "cubic",
+            "polynomial",
         ]
         for kind in kinds:
-            actual = array.resample(time="1h").interpolate(kind)
+            kwargs = {}
+            if kind == "polynomial":
+                kwargs["order"] = 1
+            actual = array.resample(time="1h").interpolate(kind, **kwargs)
+            # using interp1d, polynomial order is to set directly in kind using int
             f = interp1d(
                 np.arange(len(times)),
                 data,
-                kind=kind,
+                kind=kwargs["order"] if kind == "polynomial" else kind,
                 axis=-1,
                 bounds_error=True,
                 assume_sorted=True,
@@ -2147,14 +2152,19 @@ def test_upsample_interpolate_dask(self, chunked_time: bool) -> None:
             "slinear",
             "quadratic",
             "cubic",
+            "polynomial",
         ]
         for kind in kinds:
-            actual = array.chunk(chunks).resample(time="1h").interpolate(kind)
+            kwargs = {}
+            if kind == "polynomial":
+                kwargs["order"] = 1
+            actual = array.chunk(chunks).resample(time="1h").interpolate(kind, **kwargs)
             actual = actual.compute()
+            # using interp1d, polynomial order is to set directly in kind using int
             f = interp1d(
                 np.arange(len(times)),
                 data,
-                kind=kind,
+                kind=kwargs["order"] if kind == "polynomial" else kind,
                 axis=-1,
                 bounds_error=True,
                 assume_sorted=True,