From 516307e5f7ae33d2588f6814d612979bd6ee7b43 Mon Sep 17 00:00:00 2001
From: Khanna <avkhann2@fa18-cs425-g05-01.cs.illinois.edu>
Date: Sun, 2 Dec 2018 20:31:04 -0600
Subject: [PATCH] Spark Apps Done
---
apache-spark/python/join.py | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/apache-spark/python/join.py b/apache-spark/python/join.py
index f457116..220369d 100755
--- a/apache-spark/python/join.py
+++ b/apache-spark/python/join.py
@@ -34,12 +34,12 @@ if __name__ == "__main__":
lines = spark.read.text(sys.argv[1]).rdd.map(lambda r: r[0])
lines = lines.map(lambda a: a.encode('utf-8'))
- one = lines.filter(lambda x: x if 'profileName' in x else '')
- names = one.map(lambda y: y.split(':')[1])
-
+ names = lines.filter(lambda x: x.split(':')[1] if 'profileName' in x else '')
profiles = names.map(lambda c: (c, random.randint(1, 10)))
+
#randos = randos.map(lambda d: (d, random.randint(1, 10))
#final = profiles.join(randos)
+
res = profiles.collect()
for r in res:
--
GitLab