From dc9fd7a3adc863510d79a718e919b6833a86729c Mon Sep 17 00:00:00 2001
From: Patrick Simianer <p@simianer.de>
Date: Fri, 23 Sep 2011 20:53:15 +0200
Subject: begin refactoring

---
 dtrain/hstreaming/avgweights.rb | 27 +++++++++++++++++++++++++++
 1 file changed, 27 insertions(+)
 create mode 100755 dtrain/hstreaming/avgweights.rb

(limited to 'dtrain/hstreaming/avgweights.rb')

diff --git a/dtrain/hstreaming/avgweights.rb b/dtrain/hstreaming/avgweights.rb
new file mode 100755
index 00000000..d5cfaa4d
--- /dev/null
+++ b/dtrain/hstreaming/avgweights.rb
@@ -0,0 +1,27 @@
+#!/usr/bin/env ruby1.9.1
+
+
+STDIN.set_encoding 'utf-8'
+
+#shard_count_key = "__SHARD_COUNT__"
+
+w = {}
+c = {}
+w.default = 0
+c.default = 0
+while line = STDIN.gets
+  key, val = line.split /\t/
+  w[key] += val.to_f
+  c[key] += 1.0
+end
+
+#shard_count = w["__SHARD_COUNT__"]
+
+w.each_key { |k|
+  #if k == shard_count_key then next end
+  #if k == "__bias" then next end
+  puts "#{k}\t#{w[k]/c[k]}" #{w[k]/shard_count}"
+}
+
+#puts "#{shard_count_key}\t#{w[shard_count_key]}"
+
-- 
cgit v1.2.3