Remvoe pandas 2 checks (#1100)

Author: phofl

dask_expr/_reductions.py

@@ -7,7 +7,6 @@
 from dask.array import chunk
 from dask.array.reductions import moment_agg, moment_chunk, moment_combine, nannumel
 from dask.dataframe import hyperloglog, methods
-from dask.dataframe._compat import PANDAS_GE_200
 from dask.dataframe.core import (
     _concat,
     _cov_corr_agg,
@@ -633,7 +632,7 @@ def chunk_kwargs(self):
         out = {"keep": self.keep}
         if is_dataframe_like(self.frame._meta):
             out["subset"] = self.subset
-        if PANDAS_GE_200 and not is_index_like(self.frame._meta):
+        if not is_index_like(self.frame._meta):
             out["ignore_index"] = self.ignore_index
         return out
 

dask_expr/tests/test_datasets.py

@@ -2,7 +2,6 @@
 import sys
 
 import pytest
-from dask.dataframe._compat import PANDAS_GE_200
 
 from dask_expr import new_collection
 from dask_expr._expr import Lengths
@@ -124,7 +123,7 @@ def test_timeseries_gaph_size(seed):
     graph_size = sys.getsizeof(pickle.dumps(df.dask))
     graph_size_dd = sys.getsizeof(pickle.dumps(dict(ddf.dask)))
     # Make sure we are close to the dask.dataframe graph size
-    threshold = 1.10 if PANDAS_GE_200 else 1.50
+    threshold = 1.10
     assert graph_size < threshold * graph_size_dd
 
 

dask_expr/tests/test_map_partitions_overlap.py

@@ -2,7 +2,6 @@
 
 import numpy as np
 import pytest
-from dask.dataframe._compat import PANDAS_GE_200
 
 from dask_expr import from_pandas, map_overlap, map_partitions
 from dask_expr.tests._util import _backend_library, assert_eq
@@ -64,8 +63,6 @@ def test_map_partitions_merge(opt):
     # Check result with/without fusion
     expect = pdf1.merge(pdf2, on="x")
     df3 = (df3.optimize() if opt else df3)[list(expect.columns)]
-    if not PANDAS_GE_200:
-        df3 = df3.reset_index(drop=True)
     assert_eq(df3, expect, check_index=False)
 
 

dask_expr/tests/test_string_accessor.py

@@ -1,5 +1,4 @@
 import pytest
-from dask.dataframe._compat import PANDAS_GE_200
 
 from dask_expr._collection import DataFrame, from_pandas
 from dask_expr.tests._util import _backend_library, assert_eq
@@ -28,14 +27,7 @@ def dser(ser):
         ("contains", {"pat": "a"}),
         ("count", {"pat": "a"}),
         ("endswith", {"pat": "a"}),
-        pytest.param(
-            "extract",
-            {"pat": r"[ab](\d)"},
-            marks=pytest.mark.skipif(
-                not PANDAS_GE_200,
-                reason="Index metadata wrong for pandas<2.0",
-            ),
-        ),
+        ("extract", {"pat": r"[ab](\d)"}),
         ("extractall", {"pat": r"[ab](\d)"}),
         ("find", {"sub": "a"}),
         ("findall", {"pat": "a"}),