[DOC] Updated the setup instructions for macos and windows
Path for Spark and Hadoop to navigate some common problems
around .exe execution privileges.
Closes #65.
diff --git a/_src/install-systemml.html b/_src/install-systemml.html
index 6665363..03f23c6 100644
--- a/_src/install-systemml.html
+++ b/_src/install-systemml.html
@@ -54,11 +54,69 @@
<p class="indent">Apache Spark 2.x</p>
<p class="indent">Set SPARK_HOME to a location where Spark 2.x is installed.</p>
+ <div id="prerequisite-tabs">
+ <ul>
+ <li><a href="#prerequisite-tabs-1">MacOS/Linux</a></li>
+ <li><a href="#prerequisite-tabs-2">Windows</a></li>
+ </ul>
+
+ <div id="prerequisite-tabs-1">
+ 1) Java <br />
+ Make sure Java version is >= 1.8 and JAVA_HOME environment variable is set:
+ {% highlight bash %}
+java -version
+export JAVA_HOME="$(/usr/libexec/java_home)"{% endhighlight %}
+
+ 2) Spark <br />
+ Download Spark from <a href="https://spark.apache.org/downloads.html">https://spark.apache.org/downloads.html</a> and move to home directory, and extract. Also, set environment variables to point to the extracted directory
+ {% highlight bash %}
+export SPARK_HOME="$HOME/spark-2.1.0-bin-hadoop2.7"
+export HADOOP_HOME=$SPARK_HOME
+export SPARK_LOCAL_IP=127.0.0.1{% endhighlight %}
+
+ 3) Python and Jupyter <br />
+ Download and install Anaconda Python 3+ from <a href="https://www.anaconda.com/distribution/#download-section">https://www.anaconda.com/distribution/#download-section</a> (includes jupyter, and pip)
+ {% highlight bash %}
+export PYSPARK_DRIVER_PYTHON=jupyter
+export PYSPARK_DRIVER_PYTHON_OPTS='notebook' $SPARK_HOME/bin/pyspark --master local[*] --driver-memory 8G{% endhighlight %}
+ </div>
+
+ <div id="prerequisite-tabs-2">
+ 1) Java <br />
+ Make sure Java version is >= 1.8. Also, set JAVA_HOME environment variable and include %JAVA_HOME%\bin in the environment variable PATH:
+ {% highlight bash %}
+java -version
+ls "%JAVA_HOME%"{% endhighlight %}
+
+ 2) Spark <br />
+ Download Spark from <a href="https://spark.apache.org/downloads.html">https://spark.apache.org/downloads.html</a> and extract. Set the environment variable SPARK_HOME to point to the extracted directory. <br />
+
+ 3) Install winutils <br />
+- Download winutils.exe from <a href="http://github.com/steveloughran/winutils/raw/master/hadoop-2.6.0/bin/winutils.exe">http://github.com/steveloughran/winutils/raw/master/hadoop-2.6.0/bin/winutils.exe</a> <br />
+- Place it in c:\winutils\bin <br />
+- Set environment variable HADOOP_HOME to point to c:\winutils <br />
+- Add c:\winutils\bin to the environment variable PATH. <br />
+- Finally, modify permission of hive directory that will be used by Spark and check if Spark is correctly installed:
+
+ {% highlight bash %}
+winutils.exe chmod 777 /tmp/hive
+%SPARK_HOME%\bin\spark-shell
+%SPARK_HOME%\bin\pyspark --master local[*] --driver-memory 8G{% endhighlight %}
+
+ 3) Python and Jupyter <br />
+ Download and install Anaconda Python 3+ from <a href="https://www.anaconda.com/distribution/#download-section">https://www.anaconda.com/distribution/#download-section</a> (includes jupyter, and pip)
+ {% highlight bash %}
+set PYSPARK_DRIVER_PYTHON=jupyter
+set PYSPARK_DRIVER_PYTHON_OPTS=notebook
+%SPARK_HOME%\bin\pyspark --master local[*] --driver-memory 8G{% endhighlight %}
+ </div>
+
+ </div>
</div>
<!-- Step 2 -->
<div class="col col-12">
- <h3><span class="circle">2</span>Setup</h3>
+ <h3><span class="circle">2</span>Setup SystemML</h3>
</div>
<div id="setup-tabs">
