From 0ee9ac0084730591435c86feb691b73c98981514 Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Wed, 4 Dec 2024 13:31:43 -0800
Subject: [PATCH 01/25] Replaces uses of cudf._lib.Column.from_unique_ptr with
 pylibcudf.Column.from_libcudf

---
 python/cudf/cudf/_lib/strings_udf.pyx |  8 ++++----
 python/cudf/cudf/_lib/utils.pyx       | 11 ++++++-----
 2 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/python/cudf/cudf/_lib/strings_udf.pyx b/python/cudf/cudf/_lib/strings_udf.pyx
index dd2fafbe07f..83f0cb850a5 100644
--- a/python/cudf/cudf/_lib/strings_udf.pyx
+++ b/python/cudf/cudf/_lib/strings_udf.pyx
@@ -1,7 +1,6 @@
 # Copyright (c) 2022-2024, NVIDIA CORPORATION.
 
 from libc.stdint cimport uint8_t, uint16_t, uintptr_t
-
 from pylibcudf.libcudf.strings_udf cimport (
     get_character_cases_table as cpp_get_character_cases_table,
     get_character_flags_table as cpp_get_character_flags_table,
@@ -27,6 +26,7 @@ from rmm.librmm.device_buffer cimport device_buffer
 from rmm.pylibrmm.device_buffer cimport DeviceBuffer
 
 from cudf._lib.column cimport Column
+from pylibcudf cimport Column as plc_Column
 
 
 def get_cuda_build_version():
@@ -52,9 +52,9 @@ def column_from_udf_string_array(DeviceBuffer d_buffer):
         c_result = move(cpp_column_from_udf_string_array(data, size))
         cpp_free_udf_string_array(data, size)
 
-    result = Column.from_unique_ptr(move(c_result))
-
-    return result
+    return Column.from_pylibcudf(
+        plc_Column.from_libcudf(move(c_result))
+    )
 
 
 def get_character_flags_table_ptr():
diff --git a/python/cudf/cudf/_lib/utils.pyx b/python/cudf/cudf/_lib/utils.pyx
index 6b3f10e1806..ff032656f80 100644
--- a/python/cudf/cudf/_lib/utils.pyx
+++ b/python/cudf/cudf/_lib/utils.pyx
@@ -16,7 +16,7 @@ from pylibcudf.libcudf.table.table_view cimport table_view
 from pylibcudf.libcudf.types cimport size_type
 
 from cudf._lib.column cimport Column
-
+from pylibcudf cimport Column as plc_Column
 try:
     import ujson as json
 except ImportError:
@@ -223,10 +223,11 @@ cdef columns_from_unique_ptr(
 
     cdef size_t i
 
-    columns = [Column.from_unique_ptr(move(dereference(it+i)))
-               for i in range(c_columns.size())]
-
-    return columns
+    return [
+        Column.from_pylibcudf(
+            plc_Column.from_libcudf(move(dereference(it+i)))
+        ) for i in range(c_columns.size())
+    ]
 
 
 cpdef columns_from_pylibcudf_table(tbl):

From 76315b0a83fc31210085aadd3c81b61586c86be0 Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Wed, 4 Dec 2024 15:15:54 -0800
Subject: [PATCH 02/25] migrate changes from Propagate failures in pandas
 integration tests

---
 .../third-party-integration/test.sh              | 16 +++++++++++-----
 .../tests/test_stumpy_distributed.py             |  2 +-
 2 files changed, 12 insertions(+), 6 deletions(-)

diff --git a/ci/cudf_pandas_scripts/third-party-integration/test.sh b/ci/cudf_pandas_scripts/third-party-integration/test.sh
index f8ddbaba0f3..30e3ffc9a43 100755
--- a/ci/cudf_pandas_scripts/third-party-integration/test.sh
+++ b/ci/cudf_pandas_scripts/third-party-integration/test.sh
@@ -26,6 +26,8 @@ main() {
     LIBS=${LIBS#[}
     LIBS=${LIBS%]}
 
+    ANY_FAILURES=0
+
     for lib in ${LIBS//,/ }; do
         lib=$(echo "$lib" | tr -d '""')
         echo "Running tests for library $lib"
@@ -56,10 +58,6 @@ main() {
         rapids-logger "Check GPU usage"
         nvidia-smi
 
-        EXITCODE=0
-        trap "EXITCODE=1" ERR
-        set +e
-
         rapids-logger "pytest ${lib}"
 
         NUM_PROCESSES=8
@@ -72,12 +70,20 @@ main() {
             fi
         done
 
+        EXITCODE=0
+        trap "EXITCODE=1" ERR
+        set +e
+
         TEST_DIR=${TEST_DIR} NUM_PROCESSES=${NUM_PROCESSES} ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh ${lib}
 
+        set -e
         rapids-logger "Test script exiting with value: ${EXITCODE}"
+        if [[ ${EXITCODE} != 0 ]]; then
+            ANY_FAILURES=1
+        fi
     done
 
-    exit ${EXITCODE}
+    exit ${ANY_FAILURES}
 }
 
 main "$@"
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_stumpy_distributed.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_stumpy_distributed.py
index 0777d982ac2..f275659288e 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_stumpy_distributed.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_stumpy_distributed.py
@@ -41,7 +41,7 @@ def test_multidimensional_distributed_timeseries(dask_client):
     rng = np.random.default_rng(seed=42)
     # Each row represents data from a different dimension while each column represents
     # data from the same dimension
-    your_time_series = rng.random(3, 1000)
+    your_time_series = rng.random((3, 1000))
     # Approximately, how many data points might be found in a pattern
     window_size = 50
 

From 719da3af09cf345526c68b5bde4528027ec2801c Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Wed, 4 Dec 2024 15:20:27 -0800
Subject: [PATCH 03/25] clean up

---
 python/cudf/cudf/_lib/strings_udf.pyx |  7 +++----
 python/cudf/cudf/_lib/utils.pyx       | 11 +++++------
 2 files changed, 8 insertions(+), 10 deletions(-)

diff --git a/python/cudf/cudf/_lib/strings_udf.pyx b/python/cudf/cudf/_lib/strings_udf.pyx
index 83f0cb850a5..b77ca0c1ce5 100644
--- a/python/cudf/cudf/_lib/strings_udf.pyx
+++ b/python/cudf/cudf/_lib/strings_udf.pyx
@@ -26,7 +26,6 @@ from rmm.librmm.device_buffer cimport device_buffer
 from rmm.pylibrmm.device_buffer cimport DeviceBuffer
 
 from cudf._lib.column cimport Column
-from pylibcudf cimport Column as plc_Column
 
 
 def get_cuda_build_version():
@@ -52,9 +51,9 @@ def column_from_udf_string_array(DeviceBuffer d_buffer):
         c_result = move(cpp_column_from_udf_string_array(data, size))
         cpp_free_udf_string_array(data, size)
 
-    return Column.from_pylibcudf(
-        plc_Column.from_libcudf(move(c_result))
-    )
+    result = Column.from_unique_ptr(move(c_result))
+
+    return result
 
 
 def get_character_flags_table_ptr():
diff --git a/python/cudf/cudf/_lib/utils.pyx b/python/cudf/cudf/_lib/utils.pyx
index ff032656f80..6b3f10e1806 100644
--- a/python/cudf/cudf/_lib/utils.pyx
+++ b/python/cudf/cudf/_lib/utils.pyx
@@ -16,7 +16,7 @@ from pylibcudf.libcudf.table.table_view cimport table_view
 from pylibcudf.libcudf.types cimport size_type
 
 from cudf._lib.column cimport Column
-from pylibcudf cimport Column as plc_Column
+
 try:
     import ujson as json
 except ImportError:
@@ -223,11 +223,10 @@ cdef columns_from_unique_ptr(
 
     cdef size_t i
 
-    return [
-        Column.from_pylibcudf(
-            plc_Column.from_libcudf(move(dereference(it+i)))
-        ) for i in range(c_columns.size())
-    ]
+    columns = [Column.from_unique_ptr(move(dereference(it+i)))
+               for i in range(c_columns.size())]
+
+    return columns
 
 
 cpdef columns_from_pylibcudf_table(tbl):

From 29798e96b008457070ac5ee24c3c20545e450df5 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Thu, 5 Dec 2024 18:48:51 -0500
Subject: [PATCH 04/25] remove deleted line

---
 python/cudf/cudf/_lib/strings_udf.pyx | 1 +
 1 file changed, 1 insertion(+)

diff --git a/python/cudf/cudf/_lib/strings_udf.pyx b/python/cudf/cudf/_lib/strings_udf.pyx
index b77ca0c1ce5..dd2fafbe07f 100644
--- a/python/cudf/cudf/_lib/strings_udf.pyx
+++ b/python/cudf/cudf/_lib/strings_udf.pyx
@@ -1,6 +1,7 @@
 # Copyright (c) 2022-2024, NVIDIA CORPORATION.
 
 from libc.stdint cimport uint8_t, uint16_t, uintptr_t
+
 from pylibcudf.libcudf.strings_udf cimport (
     get_character_cases_table as cpp_get_character_cases_table,
     get_character_flags_table as cpp_get_character_flags_table,

From 5b870f49ab1d458d3699963951182c468e00fd8f Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Fri, 6 Dec 2024 04:45:29 -0800
Subject: [PATCH 05/25] add to ci job

---
 .github/workflows/pr.yaml | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/.github/workflows/pr.yaml b/.github/workflows/pr.yaml
index 7c0bd6d52e2..fe309852a45 100644
--- a/.github/workflows/pr.yaml
+++ b/.github/workflows/pr.yaml
@@ -40,6 +40,7 @@ jobs:
       - pandas-tests
       - pandas-tests-diff
       - telemetry-setup
+      - third-party-integration-tests-cudf-pandas
     secrets: inherit
     uses: rapidsai/shared-workflows/.github/workflows/pr-builder.yaml@branch-25.02
     if: always()
@@ -325,6 +326,17 @@ jobs:
         node_type: cpu4
         build_type: pull-request
         run_script: "ci/cudf_pandas_scripts/pandas-tests/diff.sh"
+  third-party-integration-tests-cudf-pandas:
+    needs: wheel-build-cudf
+    secrets: inherit
+    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@branch-24.10
+    with:
+      build_type: pull-request
+      node_type: "gpu-v100-latest-1"
+      arch: "amd64"
+      container_image: "rapidsai/ci-conda:latest"
+      run_script: |
+        ci/cudf_pandas_scripts/third-party-integration/test.sh python/cudf/cudf_pandas_tests/third_party_integration_tests/dependencies.yaml
 
   telemetry-summarize:
     runs-on: ubuntu-latest

From 30a9391a6c18da5a5c39b992dc12ee83a715888e Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Fri, 6 Dec 2024 12:20:44 -0800
Subject: [PATCH 06/25] xfail failing tests

---
 .../third_party_integration_tests/tests/test_holoviews.py      | 3 +++
 .../third_party_integration_tests/tests/test_matplotlib.py     | 3 +++
 .../third_party_integration_tests/tests/test_numpy.py          | 3 +++
 .../third_party_integration_tests/tests/test_pytorch.py        | 3 +++
 .../third_party_integration_tests/tests/test_seaborn.py        | 3 +++
 .../third_party_integration_tests/tests/test_tensorflow.py     | 1 +
 .../third_party_integration_tests/tests/test_xgboost.py        | 3 +++
 7 files changed, 19 insertions(+)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
index bef02c86355..7ac26deb673 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
@@ -75,5 +75,8 @@ def test_holoviews_histogram(df):
     return get_plot_info(hv.Histogram(df.values))
 
 
+@pytest.mark.xfail(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_holoviews_hexbin(df):
     return get_plot_info(hv.HexTiles(df, kdims=["x", "y"], vdims="y"))
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py
index 1909392b9f7..6dc23017b38 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py
@@ -33,6 +33,9 @@ def assert_plots_equal(expect, got):
 pytestmark = pytest.mark.assert_eq(fn=assert_plots_equal)
 
 
+@pytest.mark.xfail(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_line():
     df = pd.DataFrame({"x": [1, 2, 3, 4, 5], "y": [2, 4, 6, 8, 10]})
     (data,) = plt.plot(df["x"], df["y"], marker="o", linestyle="-")
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py
index 472f1889354..3f332e6fd60 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py
@@ -37,6 +37,9 @@ def test_numpy_dot(df):
     return np.dot(df, df.T)
 
 
+@pytest.mark.xfail(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_numpy_fft(sr):
     fft = np.fft.fft(sr)
     return fft
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py
index ad287471aa0..57655547161 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py
@@ -116,6 +116,9 @@ def test_torch_train(data):
     return model(test_x1, test_x2)
 
 
+@pytest.mark.xfail(
+    reason="AssertionError: The values for attribute 'device' do not match: cpu != cuda:0."
+)
 def test_torch_tensor_ctor():
     s = pd.Series(range(5))
     return torch.tensor(s.values)
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py
index 021c5bac9b7..b1a5bc6c3c4 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py
@@ -54,6 +54,9 @@ def test_scatter(df):
     return ax
 
 
+@pytest.mark.xfail(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_lineplot_with_sns_data():
     df = sns.load_dataset("flights")
     ax = sns.lineplot(data=df, x="month", y="passengers")
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py
index ba1f518cbfd..b4fad3024e7 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py
@@ -271,6 +271,7 @@ def call(self, values):
         return tf.concat(values, axis=-1)
 
 
+@pytest.mark.xfail(reason="ValueError: Invalid dtype: object")
 def test_full_example_train_with_df(df, target):
     # https://www.tensorflow.org/tutorials/load_data/pandas_dataframe#full_example
     # Inputs are directly passed as dictionary of series
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
index 70f1e6a4250..fb8de793290 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
@@ -113,6 +113,9 @@ def test_with_external_memory(
     return predt
 
 
+@pytest.mark.xfail(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 @pytest.mark.parametrize("device", ["cpu", "cuda"])
 def test_predict(device: str) -> np.ndarray:
     reg = xgb.XGBRegressor(n_estimators=2, device=device)

From ce69e269726a7e13fbc3d656ffdd8573c1755a4e Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Fri, 6 Dec 2024 19:31:33 -0800
Subject: [PATCH 07/25] remove xdist worksteal strategy

---
 .../third-party-integration/run-library-tests.sh              | 4 ----
 .../third_party_integration_tests/tests/test_holoviews.py     | 3 +++
 2 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
index d44d25d658c..761adaead28 100755
--- a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
+++ b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
@@ -23,8 +23,6 @@ runtest() {
     -v \
     --continue-on-collection-errors \
     --cache-clear \
-    --numprocesses=${NUM_PROCESSES} \
-    --dist=worksteal \
     ${TEST_DIR}/test_${lib}*.py
 }
 
@@ -42,8 +40,6 @@ main() {
     -v \
     --continue-on-collection-errors \
     --cache-clear \
-    --numprocesses=${NUM_PROCESSES} \
-    --dist=worksteal \
     ${TEST_DIR}/test_${lib}*.py
 }
 
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
index 7ac26deb673..4310f233fd0 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
@@ -71,6 +71,9 @@ def test_holoviews_heatmap(df):
     )
 
 
+@pytest.mark.xfail(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_holoviews_histogram(df):
     return get_plot_info(hv.Histogram(df.values))
 

From 4c327b25eaa6a47a09c4bc4f3932b6179deef929 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Sat, 7 Dec 2024 09:03:42 -0500
Subject: [PATCH 08/25] Update
 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py

---
 .../third_party_integration_tests/tests/test_xgboost.py        | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
index fb8de793290..70f1e6a4250 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
@@ -113,9 +113,6 @@ def test_with_external_memory(
     return predt
 
 
-@pytest.mark.xfail(
-    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
-)
 @pytest.mark.parametrize("device", ["cpu", "cuda"])
 def test_predict(device: str) -> np.ndarray:
     reg = xgb.XGBRegressor(n_estimators=2, device=device)

From 581f938fb26dcbc5b4361eceee63ac70641b49f9 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Sat, 7 Dec 2024 09:03:48 -0500
Subject: [PATCH 09/25] Update
 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py

---
 .../third_party_integration_tests/tests/test_tensorflow.py       | 1 -
 1 file changed, 1 deletion(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py
index b4fad3024e7..ba1f518cbfd 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py
@@ -271,7 +271,6 @@ def call(self, values):
         return tf.concat(values, axis=-1)
 
 
-@pytest.mark.xfail(reason="ValueError: Invalid dtype: object")
 def test_full_example_train_with_df(df, target):
     # https://www.tensorflow.org/tutorials/load_data/pandas_dataframe#full_example
     # Inputs are directly passed as dictionary of series

From 688a5618d416f40119f01c0768e0a73450d45574 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Sat, 7 Dec 2024 09:03:55 -0500
Subject: [PATCH 10/25] Update
 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py

---
 .../third_party_integration_tests/tests/test_seaborn.py        | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py
index b1a5bc6c3c4..021c5bac9b7 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py
@@ -54,9 +54,6 @@ def test_scatter(df):
     return ax
 
 
-@pytest.mark.xfail(
-    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
-)
 def test_lineplot_with_sns_data():
     df = sns.load_dataset("flights")
     ax = sns.lineplot(data=df, x="month", y="passengers")

From a47fb4fd8257d27b898b4389a0e89b23e0959737 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Sat, 7 Dec 2024 09:03:59 -0500
Subject: [PATCH 11/25] Update
 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py

---
 .../third_party_integration_tests/tests/test_pytorch.py        | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py
index 57655547161..ad287471aa0 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py
@@ -116,9 +116,6 @@ def test_torch_train(data):
     return model(test_x1, test_x2)
 
 
-@pytest.mark.xfail(
-    reason="AssertionError: The values for attribute 'device' do not match: cpu != cuda:0."
-)
 def test_torch_tensor_ctor():
     s = pd.Series(range(5))
     return torch.tensor(s.values)

From dcd54e475f09dcfa9fe9be823581c819e70eeca0 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Sat, 7 Dec 2024 09:04:04 -0500
Subject: [PATCH 12/25] Update
 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py

---
 .../third_party_integration_tests/tests/test_numpy.py          | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py
index 3f332e6fd60..472f1889354 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py
@@ -37,9 +37,6 @@ def test_numpy_dot(df):
     return np.dot(df, df.T)
 
 
-@pytest.mark.xfail(
-    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
-)
 def test_numpy_fft(sr):
     fft = np.fft.fft(sr)
     return fft

From 8090316324f1d028da443cc56150ffe1cf1442af Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Sat, 7 Dec 2024 09:04:09 -0500
Subject: [PATCH 13/25] Update
 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py

---
 .../third_party_integration_tests/tests/test_matplotlib.py     | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py
index 6dc23017b38..1909392b9f7 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py
@@ -33,9 +33,6 @@ def assert_plots_equal(expect, got):
 pytestmark = pytest.mark.assert_eq(fn=assert_plots_equal)
 
 
-@pytest.mark.xfail(
-    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
-)
 def test_line():
     df = pd.DataFrame({"x": [1, 2, 3, 4, 5], "y": [2, 4, 6, 8, 10]})
     (data,) = plt.plot(df["x"], df["y"], marker="o", linestyle="-")

From 22cc7089241867ed57a412aab5bd7e3b4393318e Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Sat, 7 Dec 2024 09:04:23 -0500
Subject: [PATCH 14/25] Update
 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py

---
 .../third_party_integration_tests/tests/test_holoviews.py      | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
index 4310f233fd0..4ba906aeeba 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
@@ -78,8 +78,5 @@ def test_holoviews_histogram(df):
     return get_plot_info(hv.Histogram(df.values))
 
 
-@pytest.mark.xfail(
-    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
-)
 def test_holoviews_hexbin(df):
     return get_plot_info(hv.HexTiles(df, kdims=["x", "y"], vdims="y"))

From 725bcf321cdc2806d519d6d9699c2a835807b0c0 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Sat, 7 Dec 2024 09:04:29 -0500
Subject: [PATCH 15/25] Update
 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py

---
 .../third_party_integration_tests/tests/test_holoviews.py      | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
index 4ba906aeeba..bef02c86355 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
@@ -71,9 +71,6 @@ def test_holoviews_heatmap(df):
     )
 
 
-@pytest.mark.xfail(
-    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
-)
 def test_holoviews_histogram(df):
     return get_plot_info(hv.Histogram(df.values))
 

From 5106ab5953ed479d5dd50d47ddface7f8b4fa8d2 Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Mon, 9 Dec 2024 07:33:23 -0800
Subject: [PATCH 16/25] fix a bug and xfail a test

---
 .../third_party_integration_tests/tests/test_tensorflow.py      | 1 +
 .../third_party_integration_tests/tests/test_xgboost.py         | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py
index ba1f518cbfd..b4fad3024e7 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_tensorflow.py
@@ -271,6 +271,7 @@ def call(self, values):
         return tf.concat(values, axis=-1)
 
 
+@pytest.mark.xfail(reason="ValueError: Invalid dtype: object")
 def test_full_example_train_with_df(df, target):
     # https://www.tensorflow.org/tutorials/load_data/pandas_dataframe#full_example
     # Inputs are directly passed as dictionary of series
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
index 70f1e6a4250..4f2556f12a8 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
@@ -123,7 +123,7 @@ def test_predict(device: str) -> np.ndarray:
 
     predt0 = reg.predict(X_df)
 
-    predt1 = booster.inplace_predict(X_df)
+    predt1 = booster.inplace_predict(X_df).get()
     np.testing.assert_allclose(predt0, predt1)
 
     predt2 = booster.predict(xgb.DMatrix(X_df))

From 75641d51fa79f883705a60d2f76ea02b0f0f3f66 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Mon, 9 Dec 2024 12:30:07 -0500
Subject: [PATCH 17/25] remove get call

---
 .../third_party_integration_tests/tests/test_xgboost.py         | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
index 4f2556f12a8..70f1e6a4250 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
@@ -123,7 +123,7 @@ def test_predict(device: str) -> np.ndarray:
 
     predt0 = reg.predict(X_df)
 
-    predt1 = booster.inplace_predict(X_df).get()
+    predt1 = booster.inplace_predict(X_df)
     np.testing.assert_allclose(predt0, predt1)
 
     predt2 = booster.predict(xgb.DMatrix(X_df))

From 5b11e6671b02dbda18d282b3a686b3fb14c30887 Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Wed, 11 Dec 2024 07:55:04 -0800
Subject: [PATCH 18/25] skip tests

---
 .../third-party-integration/run-library-tests.sh            | 6 ++++++
 .../third_party_integration_tests/tests/test_holoviews.py   | 3 +++
 .../third_party_integration_tests/tests/test_matplotlib.py  | 6 ++++++
 .../third_party_integration_tests/tests/test_numpy.py       | 3 +++
 .../third_party_integration_tests/tests/test_pytorch.py     | 3 +++
 .../third_party_integration_tests/tests/test_seaborn.py     | 3 +++
 .../third_party_integration_tests/tests/test_xgboost.py     | 3 +++
 7 files changed, 27 insertions(+)

diff --git a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
index 761adaead28..7f2ab32412c 100755
--- a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
+++ b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
@@ -20,9 +20,12 @@ runtest() {
 
     pytest \
     $plugin \
+    --ignore=${TEST_DIR}/test_catboost.py \
     -v \
     --continue-on-collection-errors \
     --cache-clear \
+    --numprocesses=${NUM_PROCESSES} \
+    --dist=worksteal \
     ${TEST_DIR}/test_${lib}*.py
 }
 
@@ -37,9 +40,12 @@ main() {
     pytest \
     --compare \
     -p cudf.pandas \
+    --ignore=${TEST_DIR}/test_catboost.py \
     -v \
     --continue-on-collection-errors \
     --cache-clear \
+    --numprocesses=${NUM_PROCESSES} \
+    --dist=worksteal \
     ${TEST_DIR}/test_${lib}*.py
 }
 
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
index bef02c86355..8be48953974 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_holoviews.py
@@ -71,6 +71,9 @@ def test_holoviews_heatmap(df):
     )
 
 
+@pytest.mark.skip(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_holoviews_histogram(df):
     return get_plot_info(hv.Histogram(df.values))
 
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py
index 1909392b9f7..c91808021e8 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_matplotlib.py
@@ -33,6 +33,9 @@ def assert_plots_equal(expect, got):
 pytestmark = pytest.mark.assert_eq(fn=assert_plots_equal)
 
 
+@pytest.mark.skip(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_line():
     df = pd.DataFrame({"x": [1, 2, 3, 4, 5], "y": [2, 4, 6, 8, 10]})
     (data,) = plt.plot(df["x"], df["y"], marker="o", linestyle="-")
@@ -40,6 +43,9 @@ def test_line():
     return plt.gca()
 
 
+@pytest.mark.skip(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_bar():
     data = pd.Series([1, 2, 3, 4, 5], index=["a", "b", "c", "d", "e"])
     ax = data.plot(kind="bar")
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py
index 472f1889354..4d35d9e8946 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_numpy.py
@@ -37,6 +37,9 @@ def test_numpy_dot(df):
     return np.dot(df, df.T)
 
 
+@pytest.mark.skip(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_numpy_fft(sr):
     fft = np.fft.fft(sr)
     return fft
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py
index ad287471aa0..7cea635afc4 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_pytorch.py
@@ -116,6 +116,9 @@ def test_torch_train(data):
     return model(test_x1, test_x2)
 
 
+@pytest.mark.skip(
+    reason="AssertionError: The values for attribute 'device' do not match: cpu != cuda:0."
+)
 def test_torch_tensor_ctor():
     s = pd.Series(range(5))
     return torch.tensor(s.values)
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py
index 021c5bac9b7..f6a8a96ae3c 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_seaborn.py
@@ -54,6 +54,9 @@ def test_scatter(df):
     return ax
 
 
+@pytest.mark.skip(
+    reason="AttributeError: 'ndarray' object has no attribute '_fsproxy_wrapped'"
+)
 def test_lineplot_with_sns_data():
     df = sns.load_dataset("flights")
     ax = sns.lineplot(data=df, x="month", y="passengers")
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
index 70f1e6a4250..0fd632507a6 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_xgboost.py
@@ -113,6 +113,9 @@ def test_with_external_memory(
     return predt
 
 
+@pytest.mark.skip(
+    reason="TypeError: Implicit conversion to a NumPy array is not allowed. Please use `.get()` to construct a NumPy array explicitly."
+)
 @pytest.mark.parametrize("device", ["cpu", "cuda"])
 def test_predict(device: str) -> np.ndarray:
     reg = xgb.XGBRegressor(n_estimators=2, device=device)

From 0b4f5fa4a210c2905146ece641f7ed00454a3a1c Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Wed, 11 Dec 2024 12:07:45 -0800
Subject: [PATCH 19/25] import or skip catboost

---
 .../third-party-integration/run-library-tests.sh          | 2 --
 .../third_party_integration_tests/tests/test_catboost.py  | 8 +++++++-
 2 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
index 7f2ab32412c..d44d25d658c 100755
--- a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
+++ b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
@@ -20,7 +20,6 @@ runtest() {
 
     pytest \
     $plugin \
-    --ignore=${TEST_DIR}/test_catboost.py \
     -v \
     --continue-on-collection-errors \
     --cache-clear \
@@ -40,7 +39,6 @@ main() {
     pytest \
     --compare \
     -p cudf.pandas \
-    --ignore=${TEST_DIR}/test_catboost.py \
     -v \
     --continue-on-collection-errors \
     --cache-clear \
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
index 04cc69231fe..a1c2b33b344 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
@@ -3,7 +3,13 @@
 import numpy as np
 import pandas as pd
 import pytest
-from catboost import CatBoostClassifier, CatBoostRegressor, Pool
+
+try:
+    from catboost import CatBoostClassifier, CatBoostRegressor, Pool
+except Exception:
+    pytest.skip(
+        "ValueError: numpy.dtype size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject"
+    )
 from sklearn.datasets import make_classification, make_regression
 
 rng = np.random.default_rng(seed=42)

From 4a2f78af57d5c6ba8b6611d7b169900c393b3886 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Wed, 11 Dec 2024 18:04:17 -0500
Subject: [PATCH 20/25] Update
 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py

---
 .../third_party_integration_tests/tests/test_catboost.py        | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
index a1c2b33b344..ca472041458 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
@@ -8,7 +8,7 @@
     from catboost import CatBoostClassifier, CatBoostRegressor, Pool
 except Exception:
     pytest.skip(
-        "ValueError: numpy.dtype size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject"
+        reason="ValueError: numpy.dtype size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject", allow_module_level=True
     )
 from sklearn.datasets import make_classification, make_regression
 

From 3d70643414d10e881ec71e1427fceb2ef37dc6d4 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Wed, 11 Dec 2024 19:08:58 -0500
Subject: [PATCH 21/25] clean up

---
 .../third_party_integration_tests/tests/test_catboost.py       | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
index ca472041458..6f9db8194c0 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
@@ -8,7 +8,8 @@
     from catboost import CatBoostClassifier, CatBoostRegressor, Pool
 except Exception:
     pytest.skip(
-        reason="ValueError: numpy.dtype size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject", allow_module_level=True
+        reason="ValueError: numpy.dtype size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject",
+        allow_module_level=True,
     )
 from sklearn.datasets import make_classification, make_regression
 

From ead0792494f88081a5010655702f64087b63b637 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Wed, 11 Dec 2024 22:06:20 -0500
Subject: [PATCH 22/25] ignore catboost tests

---
 .../third-party-integration/run-library-tests.sh                | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
index d44d25d658c..7f2ab32412c 100755
--- a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
+++ b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
@@ -20,6 +20,7 @@ runtest() {
 
     pytest \
     $plugin \
+    --ignore=${TEST_DIR}/test_catboost.py \
     -v \
     --continue-on-collection-errors \
     --cache-clear \
@@ -39,6 +40,7 @@ main() {
     pytest \
     --compare \
     -p cudf.pandas \
+    --ignore=${TEST_DIR}/test_catboost.py \
     -v \
     --continue-on-collection-errors \
     --cache-clear \

From 93d15a90e5cb178d0be5c8e43dd9259f2b667d5b Mon Sep 17 00:00:00 2001
From: Matthew Murray <matthewmurray711@gmail.com>
Date: Thu, 12 Dec 2024 06:04:00 -0800
Subject: [PATCH 23/25] remove catboost tests

---
 .../run-library-tests.sh                      |   2 -
 .../dependencies.yaml                         |  15 --
 .../tests/test_catboost.py                    | 136 ------------------
 3 files changed, 153 deletions(-)
 delete mode 100644 python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py

diff --git a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
index 7f2ab32412c..d44d25d658c 100755
--- a/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
+++ b/ci/cudf_pandas_scripts/third-party-integration/run-library-tests.sh
@@ -20,7 +20,6 @@ runtest() {
 
     pytest \
     $plugin \
-    --ignore=${TEST_DIR}/test_catboost.py \
     -v \
     --continue-on-collection-errors \
     --cache-clear \
@@ -40,7 +39,6 @@ main() {
     pytest \
     --compare \
     -p cudf.pandas \
-    --ignore=${TEST_DIR}/test_catboost.py \
     -v \
     --continue-on-collection-errors \
     --cache-clear \
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/dependencies.yaml b/python/cudf/cudf_pandas_tests/third_party_integration_tests/dependencies.yaml
index e726b7fdca1..3891110e9d3 100644
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/dependencies.yaml
+++ b/python/cudf/cudf_pandas_tests/third_party_integration_tests/dependencies.yaml
@@ -76,13 +76,6 @@ files:
       - py_version
       - test_base
       - test_xgboost
-  test_catboost:
-    output: none
-    includes:
-      - cuda_version
-      - py_version
-      - test_base
-      - test_catboost
   test_cuml:
     output: none
     includes:
@@ -251,14 +244,6 @@ dependencies:
           - pip
           - pip:
             - xgboost>=2.0.1
-  test_catboost:
-    common:
-      - output_types: conda
-        packages:
-          - numpy
-          - scipy
-          - scikit-learn
-          - catboost
   test_cuml:
     common:
       - output_types: conda
diff --git a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py b/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
deleted file mode 100644
index 6f9db8194c0..00000000000
--- a/python/cudf/cudf_pandas_tests/third_party_integration_tests/tests/test_catboost.py
+++ /dev/null
@@ -1,136 +0,0 @@
-# Copyright (c) 2024, NVIDIA CORPORATION.
-
-import numpy as np
-import pandas as pd
-import pytest
-
-try:
-    from catboost import CatBoostClassifier, CatBoostRegressor, Pool
-except Exception:
-    pytest.skip(
-        reason="ValueError: numpy.dtype size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject",
-        allow_module_level=True,
-    )
-from sklearn.datasets import make_classification, make_regression
-
-rng = np.random.default_rng(seed=42)
-
-
-def assert_catboost_equal(expect, got, rtol=1e-7, atol=0.0):
-    if isinstance(expect, (tuple, list)):
-        assert len(expect) == len(got)
-        for e, g in zip(expect, got):
-            assert_catboost_equal(e, g, rtol, atol)
-    elif isinstance(expect, np.ndarray):
-        np.testing.assert_allclose(expect, got, rtol=rtol, atol=atol)
-    elif isinstance(expect, pd.DataFrame):
-        pd.testing.assert_frame_equal(expect, got)
-    elif isinstance(expect, pd.Series):
-        pd.testing.assert_series_equal(expect, got)
-    else:
-        assert expect == got
-
-
-pytestmark = pytest.mark.assert_eq(fn=assert_catboost_equal)
-
-
-@pytest.fixture
-def regression_data():
-    X, y = make_regression(n_samples=100, n_features=10, random_state=42)
-    return pd.DataFrame(X), pd.Series(y)
-
-
-@pytest.fixture
-def classification_data():
-    X, y = make_classification(
-        n_samples=100, n_features=10, n_classes=2, random_state=42
-    )
-    return pd.DataFrame(X), pd.Series(y)
-
-
-def test_catboost_regressor_with_dataframe(regression_data):
-    X, y = regression_data
-    model = CatBoostRegressor(iterations=10, verbose=0)
-    model.fit(X, y)
-    predictions = model.predict(X)
-    return predictions
-
-
-def test_catboost_regressor_with_numpy(regression_data):
-    X, y = regression_data
-    model = CatBoostRegressor(iterations=10, verbose=0)
-    model.fit(X.values, y.values)
-    predictions = model.predict(X.values)
-    return predictions
-
-
-def test_catboost_classifier_with_dataframe(classification_data):
-    X, y = classification_data
-    model = CatBoostClassifier(iterations=10, verbose=0)
-    model.fit(X, y)
-    predictions = model.predict(X)
-    return predictions
-
-
-def test_catboost_classifier_with_numpy(classification_data):
-    X, y = classification_data
-    model = CatBoostClassifier(iterations=10, verbose=0)
-    model.fit(X.values, y.values)
-    predictions = model.predict(X.values)
-    return predictions
-
-
-def test_catboost_with_pool_and_dataframe(regression_data):
-    X, y = regression_data
-    train_pool = Pool(X, y)
-    model = CatBoostRegressor(iterations=10, verbose=0)
-    model.fit(train_pool)
-    predictions = model.predict(X)
-    return predictions
-
-
-def test_catboost_with_pool_and_numpy(regression_data):
-    X, y = regression_data
-    train_pool = Pool(X.values, y.values)
-    model = CatBoostRegressor(iterations=10, verbose=0)
-    model.fit(train_pool)
-    predictions = model.predict(X.values)
-    return predictions
-
-
-def test_catboost_with_categorical_features():
-    data = {
-        "numerical_feature": rng.standard_normal(100),
-        "categorical_feature": rng.choice(["A", "B", "C"], size=100),
-        "target": rng.integers(0, 2, size=100),
-    }
-    df = pd.DataFrame(data)
-    X = df[["numerical_feature", "categorical_feature"]]
-    y = df["target"]
-    cat_features = ["categorical_feature"]
-    model = CatBoostClassifier(
-        iterations=10, verbose=0, cat_features=cat_features
-    )
-    model.fit(X, y)
-    predictions = model.predict(X)
-    return predictions
-
-
-@pytest.mark.parametrize(
-    "X, y",
-    [
-        (
-            pd.DataFrame(rng.standard_normal((100, 5))),
-            pd.Series(rng.standard_normal(100)),
-        ),
-        (rng.standard_normal((100, 5)), rng.standard_normal(100)),
-    ],
-)
-def test_catboost_train_test_split(X, y):
-    from sklearn.model_selection import train_test_split
-
-    X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)
-    model = CatBoostRegressor(iterations=10, verbose=0)
-    model.fit(X_train, y_train)
-    predictions = model.predict(X_test)
-    return len(X_train), len(X_test), len(y_train), len(y_test), predictions

From f2e7eb0f66e96eb259d11ee8c0c3e8893fc2f41f Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Thu, 12 Dec 2024 16:32:58 -0500
Subject: [PATCH 24/25] Update .github/workflows/pr.yaml

---
 .github/workflows/pr.yaml | 1 -
 1 file changed, 1 deletion(-)

diff --git a/.github/workflows/pr.yaml b/.github/workflows/pr.yaml
index fe309852a45..737a8248bf8 100644
--- a/.github/workflows/pr.yaml
+++ b/.github/workflows/pr.yaml
@@ -40,7 +40,6 @@ jobs:
       - pandas-tests
       - pandas-tests-diff
       - telemetry-setup
-      - third-party-integration-tests-cudf-pandas
     secrets: inherit
     uses: rapidsai/shared-workflows/.github/workflows/pr-builder.yaml@branch-25.02
     if: always()

From 80ea00e9ce8f204f037caa88804c4028b536d5c5 Mon Sep 17 00:00:00 2001
From: Matthew Murray <41342305+Matt711@users.noreply.github.com>
Date: Thu, 12 Dec 2024 16:33:03 -0500
Subject: [PATCH 25/25] Update .github/workflows/pr.yaml

---
 .github/workflows/pr.yaml | 11 -----------
 1 file changed, 11 deletions(-)

diff --git a/.github/workflows/pr.yaml b/.github/workflows/pr.yaml
index 737a8248bf8..7c0bd6d52e2 100644
--- a/.github/workflows/pr.yaml
+++ b/.github/workflows/pr.yaml
@@ -325,17 +325,6 @@ jobs:
         node_type: cpu4
         build_type: pull-request
         run_script: "ci/cudf_pandas_scripts/pandas-tests/diff.sh"
-  third-party-integration-tests-cudf-pandas:
-    needs: wheel-build-cudf
-    secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@branch-24.10
-    with:
-      build_type: pull-request
-      node_type: "gpu-v100-latest-1"
-      arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
-      run_script: |
-        ci/cudf_pandas_scripts/third-party-integration/test.sh python/cudf/cudf_pandas_tests/third_party_integration_tests/dependencies.yaml
 
   telemetry-summarize:
     runs-on: ubuntu-latest