summaryrefslogtreecommitdiff
path: root/dtrain/job2/hadoop-streaming-job.sh
diff options
context:
space:
mode:
authorPatrick Simianer <p@simianer.de>2011-09-23 20:53:15 +0200
committerPatrick Simianer <p@simianer.de>2011-09-23 20:53:15 +0200
commitdc9fd7a3adc863510d79a718e919b6833a86729c (patch)
tree4baf0c6cadff000a20039994237ccaf468daee40 /dtrain/job2/hadoop-streaming-job.sh
parent1ad0eb820ee946e5a142567380fc0488c9a5d6de (diff)
begin refactoring
Diffstat (limited to 'dtrain/job2/hadoop-streaming-job.sh')
-rwxr-xr-xdtrain/job2/hadoop-streaming-job.sh23
1 files changed, 0 insertions, 23 deletions
diff --git a/dtrain/job2/hadoop-streaming-job.sh b/dtrain/job2/hadoop-streaming-job.sh
deleted file mode 100755
index 9ee70a33..00000000
--- a/dtrain/job2/hadoop-streaming-job.sh
+++ /dev/null
@@ -1,23 +0,0 @@
-#!/bin/bash
-
-HADOOP_HOME=/usr/lib/hadoop-0.20
-JAR=contrib/streaming/hadoop-streaming-0.20.2-cdh3u1.jar
-HSTREAMING="$HADOOP_HOME/bin/hadoop jar $HADOOP_HOME/$JAR"
-
-IN=in/nc-wmt11-de-en-dyer-cs-joshua.tok.lc.fixamp1.loo.psg.dtrain
-OUT=out/nc-wmt11-de-en-dyer-cs-joshua.tok.lc.fixamp1.loo.psg.dtrain-weights.50
-
-$HSTREAMING \
- -mapper "dtrain.sh" \
- -reducer "avgweights.rb" \
- -input $IN \
- -output $OUT \
- -file avgweights.rb \
- -file dtrain.sh \
- -file dtrain \
- -file dtrain.ini \
- -file cdec.ini \
- -file nc-wmt11.en.srilm.3.gz \
- -jobconf mapred.reduce.tasks=1 \
- -jobconf mapred.max.map.failures.percent=100
-