From 87b0462cb0c1c753a7d775f8472f2c49eaa5ff82 Mon Sep 17 00:00:00 2001
From: karthik <karthik@barrelsofdata.com>
Date: Thu, 9 Jul 2020 09:21:43 -0400
Subject: [PATCH] Initial commit

---
 .../com/barrelsofdata/sparkexamples/Driver.scala      | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/src/main/scala/com/barrelsofdata/sparkexamples/Driver.scala b/src/main/scala/com/barrelsofdata/sparkexamples/Driver.scala
index 16321ca..23aedb5 100644
--- a/src/main/scala/com/barrelsofdata/sparkexamples/Driver.scala
+++ b/src/main/scala/com/barrelsofdata/sparkexamples/Driver.scala
@@ -23,7 +23,16 @@ object Driver {
       .rdd.reduceByKey(_ + _)
       .toDF("word", "frequency")
 
-    wordFrequencies.write.option("header","true").csv(outputFilePath)
+//    val wordFrequencies: DataFrame = words
+//      .groupBy(col("value")).count()
+//      .toDF("word", "frequency")
+
+    wordFrequencies
+      .coalesce(1)
+      .write
+      .option("header","true")
+      .csv(outputFilePath)
+
     LOG.info(s"Result successfully written to $outputFilePath")
   }