Added instructions on installing PySpark 3.2.0

Author: pybokeh

pyspark/installing_pyspark_320.txt

@@ -0,0 +1,55 @@
+# Linux / Debian Installation Instructions
+
+# Create a separate python environment with PySpark installed
+# Spark SQL
+pip install pyspark[sql]
+# Install Pandas API on Spark and plotly
+pip install pyspark[pandas_on_spark] plotly
+# Optional if you want to create a separate jupyter kernel
+pip install ipykernel
+python -m ipykernel install --user --name your_env --display-name "Py3.9 (pyspark_dev)"
+
+# Install Java 8, scala, and git
+sudo apt install openjdk-8-jdk scala git
+
+# Install source for Spark 3.2.0
+wget https://dlcdn.apache.org/spark/spark-3.2.0/spark-3.2.0-bin-hadoop3.2.tgz
+
+# Extract Spark to /opt/spark
+sudo mkdir /opt/spark
+sudo tar -xf spark*.tgz -C /opt/spark --strip-component 1
+
+# Change permission to /opt/spark so that Spark can write inside it
+sudo chmod -R 777 /opt/spark
+
+# Edit .profile or .bashrc to add environment variables: SPARK_HOME
+export SPARK_HOME=/opt/spark
+
+# Add spark/bin and spark/sbin to the PATH environment variable
+export PATH=$PATH:$SPARK_HOME/bin:$SPARK_HOME/sbin
+
+# Then create PYSPARK_PYTHON environment that points to your Python interpreter that has PySpark installed
+export PYSPARK_PYTHON=<path_to_python_env_with_pyspark_python>
+
+# Refresh .profile or .bashrc
+source ~/.profile
+
+
+# For Windows 10
+Install Java 8 from old Sun Microsystem's site, not Oracle:  https://www.java.com/download/ie_manual.jsp
+
+# Create separate Python virtual environment comtaining PySpark, Pandas API on Spark, and Plotly
+pip install pyspark[sql]
+pip install pyspark[pandas_on_spark] plotly
+# Optional if you want to create a separate jupyter kernel
+pip install ipykernel
+python -m ipykernel install --user --name your_env --display-name "Py3.9 (pyspark_dev)"
+
+# Set SPARK_HOME and PYSPARK_PYTHON environment variables
+set SPARK_HOME=<path_to_site_packages/pyspark_folder>
+set PYSPARK_PYTHON=<path_to_python.exe>
+
+# Trick Spark into thinking that you have Hadoop installed
+Download winutils.exe from https://github.com/cdarlint/winutils, save locally to "hadoop/bin" folder and then
+set HADOOP_HOME=[path_to_hadoop_folder]
+