summaryrefslogtreecommitdiff
path: root/dtrain/hstreaming/red-avg.rb
diff options
context:
space:
mode:
authorPatrick Simianer <p@simianer.de>2011-11-28 11:34:41 +0100
committerPatrick Simianer <p@simianer.de>2011-11-28 11:34:41 +0100
commitb6a8fdc981daa3b01826a3be4c2355ee5cd61f0a (patch)
tree7565cbd3cded132a15961571f6ed820bd5e2566c /dtrain/hstreaming/red-avg.rb
parentdd9ed2de7d5f9e3e665e532d73660aa5276680df (diff)
hstreaming stuff
Diffstat (limited to 'dtrain/hstreaming/red-avg.rb')
-rwxr-xr-xdtrain/hstreaming/red-avg.rb10
1 files changed, 6 insertions, 4 deletions
diff --git a/dtrain/hstreaming/red-avg.rb b/dtrain/hstreaming/red-avg.rb
index 771f4c0e..c0b69eb4 100755
--- a/dtrain/hstreaming/red-avg.rb
+++ b/dtrain/hstreaming/red-avg.rb
@@ -1,6 +1,5 @@
#!/usr/bin/env ruby1.9.1
-
shard_count_key = "__SHARD_COUNT__"
STDIN.set_encoding 'utf-8'
@@ -8,7 +7,7 @@ STDOUT.set_encoding 'utf-8'
w = {}
c = {}
-w.default = 0
+w.default = 0.
c.default = 0
while line = STDIN.gets
key, val = line.split /\t/
@@ -19,7 +18,10 @@ end
puts "# dtrain reducer: average"
shard_count = w["__SHARD_COUNT__"]
w.each_key { |k|
- if k == shard_count_key then next end
- puts "#{k}\t#{w[k]/shard_count}"
+ if k == shard_count_key
+ puts "# shard count: #{shard_count.to_i}"
+ else
+ puts "#{k}\t#{w[k]/shard_count}\t# #{c[k]}"
+ end
}