环境准备:anaconda3、spark
1.把pyspark文件夹复制到sitepakage中
cp -r /opt/cloudera/parcels/SPARK2-2.3.0.cloudera4-1.cdh5.13.3.p0.611179/lib/spark2/python/pyspark ./
2.设置spark_home和pythonpath
export SPARK_HOME=/opt/cloudera/parcels/SPARK2-2.3.0.cloudera4-1.cdh5.13.3.p0.611179/lib/spark2
export PYTHONPATH=/usr/local/anaconda/lib/python3.7/site-packages:/opt/cloudera/parcels/SPARK2-2.3.0.cloudera4-1.cdh5.13.3.p0.611179/bin