[SPARK-19064][PYSPARK] Fix pip installing of sub components

## What changes were proposed in this pull request?

Fix instalation of mllib and ml sub components, and more eagerly cleanup cache files during test script & make-distribution.

## How was this patch tested?

Updated sanity test script to import mllib and ml sub-components.

Author: Holden Karau <holden@us.ibm.com>

Closes #16465 from holdenk/SPARK-19064-fix-pip-install-sub-components.

(cherry picked from commit 965c82d8)
Signed-off-by: Holden Karau <holden@us.ibm.com>

dev/make-distribution.sh

@@ -213,6 +213,8 @@ cp -r "$SPARK_HOME/data" "$DISTDIR"
 if [ "$MAKE_PIP" == "true" ]; then
   echo "Building python distribution package"
   pushd "$SPARK_HOME/python" > /dev/null
+  # Delete the egg info file if it exists, this can cache older setup files.
+  rm -rf pyspark.egg-info || echo "No existing egg info file, skipping deletion"
   python setup.py sdist
   popd > /dev/null
 else

dev/pip-sanity-check.py

@@ -18,6 +18,8 @@
 from __future__ import print_function
 
 from pyspark.sql import SparkSession
+from pyspark.ml.param import Params
+from pyspark.mllib.linalg import *
 import sys
 
 if __name__ == "__main__":

dev/requirements.txt

 jira==1.0.3
 PyGithub==1.26.0
 Unidecode==0.04.19
+pypandoc==1.3.3

dev/run-pip-tests

@@ -78,11 +78,14 @@ for python in "${PYTHON_EXECS[@]}"; do
     mkdir -p "$VIRTUALENV_PATH"
     virtualenv --python=$python "$VIRTUALENV_PATH"
     source "$VIRTUALENV_PATH"/bin/activate
-    # Upgrade pip
-    pip install --upgrade pip
+    # Upgrade pip & friends
+    pip install --upgrade pip pypandoc wheel
+    pip install numpy # Needed so we can verify mllib imports
 
     echo "Creating pip installable source dist"
     cd "$FWDIR"/python
+    # Delete the egg info file if it exists, this can cache the setup file.
+    rm -rf pyspark.egg-info || echo "No existing egg info file, skipping deletion"
     $python setup.py sdist
 
 

python/setup.py

@@ -162,7 +162,12 @@ try:
         url='https://github.com/apache/spark/tree/master/python',
         packages=['pyspark',
                   'pyspark.mllib',
+                  'pyspark.mllib.linalg',
+                  'pyspark.mllib.stat',
                   'pyspark.ml',
+                  'pyspark.ml.linalg',
+                  'pyspark.ml.param',
+                  'pyspark.ml.stat',
                   'pyspark.sql',
                   'pyspark.streaming',
                   'pyspark.bin',