reverting numpy to 1.19.2 to solve pyarrow integration issue, changes to resolve numpy related build issues and adding regression test

Author: Nikhil Raverkar

docker/0.23-1/final/Dockerfile.cpu

@@ -9,9 +9,11 @@ RUN python -m pip install -r /requirements.txt && \
 
 COPY dist/sagemaker_sklearn_container-2.0-py3-none-any.whl /sagemaker_sklearn_container-2.0-py3-none-any.whl
 # https://github.com/googleapis/google-cloud-python/issues/6647
-RUN rm -rf /miniconda3/lib/python3.7/site-packages/numpy-1.19.4.dist-info && \
+RUN conda install numpy --force-reinstall
+    rm -rf /miniconda3/lib/python3.7/site-packages/numpy-1.19.4.dist-info && \
     pip install --no-cache /sagemaker_sklearn_container-2.0-py3-none-any.whl && \
-    rm /sagemaker_sklearn_container-2.0-py3-none-any.whl
+    rm /sagemaker_sklearn_container-2.0-py3-none-any.whl && \
+    rm /miniconda3/lib/python3.8/site-packages/numpy-1.21.2.dist-info
 
 ENV SAGEMAKER_TRAINING_MODULE sagemaker_sklearn_container.training:main
 ENV SAGEMAKER_SERVING_MODULE sagemaker_sklearn_container.serving:main

requirements.txt

@@ -6,7 +6,7 @@ itsdangerous==2.0.1
 gunicorn==20.0.4
 model-archiver==1.0.3
 multi-model-server==1.1.1
-numpy==1.21.0
+numpy==1.19.2
 pandas==1.1.3
 psutil==5.7.2
 python-dateutil==2.8.1

test/unit/test_modules.py

@@ -1,4 +1,13 @@
+import pandas as pd
+
+
 def test_pandas_version():
     import pandas as pd
     major, minor, patch = pd.__version__.split('.')
     assert major == '1'
+
+
+def test_pyarrow_to_parquet_conversion_regression_issue_106():
+    df = pd.DataFrame({'x': [1, 2]})
+    df.to_parquet('test.parquet', engine='pyarrow')
+