mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jintao Guan <ilovexia...@gmail.com>
Subject intermediate files of RecommenderJob
Date Wed, 30 Oct 2013 17:06:43 GMT
Hello everyone,

I am learning the source code of Mahout.
Now I have a question about
org.apache.mahout.cf.taste.hadoop.item.RecommenderJob.

I have run this job in my local pseudo-mode HDFS using ml-100k as intput
data.

There are ten sub-jobs:

PreparePreferenceMatrixJob-ItemIDIndexMapper-Reducer
PreparePreferenceMatrixJob-ToItemPrefsMapper-Reducer
PreparePreferenceMatrixJob-ToItemVectorsMapper-Reducer
RowSimilarityJob-VectorNormMapper-Reducer
RowSimilarityJob-CooccurrencesMapper-Reducer
RowSimilarityJob-UnsymmetrifyMapper-Reducer
RecommenderJob-SimilarityMatrixRowWrapperMapper-Reducer
RecommenderJob-UserVectorSplitterMapper-Reducer
RecommenderJob-Mapper-Reducer
RecommenderJob-PartialMultiplyMapper-Reducer


I want to take a look at the output file of every sub-job.
And I have found the output files in the HDFS.

But the content of these output files is just a lot of unreadable code like
this:

�CF�C'�B��C�B��C�C8�C�C�C(�C
�B��C.�B��C���B��C'�B��Cg�C>�CO�C:�C
�C�CR�C3�C�C
�C�C�CVCH�C}C$yC�vC>uC]�CoC'�C�CdC�bCQaC`C2_C$YC'XCRC(�COCN�B�GCFCEC<@C)8CI2C��C

Is it because the compression mechanism of Mahout ?
Could someone tell me how and where to check the intermediate output files ?

Thank you all.

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message