Changeset 31364


Ignore:
Timestamp:
01/28/17 09:57:21 (4 years ago)
Author:
davidb
Message:

removed sample() line

File:
1 edited

Legend:

Unmodified
Added
Removed
  • other-projects/hathitrust/wcsa/extracted-features-solr/trunk/solr-ingest/src/main/java/org/hathitrust/extractedfeatures/ProcessForCatalogLangCount.java

    r31363 r31364  
    126126        JavaRDD<Text> json_text_rdd = input_pair_rdd.map(item -> item._2);
    127127       
    128         JavaRDD<Text> json_text_sample_rdd = json_text_rdd.sample(false,0.0001);
     128        //JavaRDD<Text> json_text_sample_rdd = json_text_rdd.sample(false,0.0001);
    129129       
    130130        /*
     
    138138        PerVolumeCatalogLangSequenceFileMap volume_catalog_langfreq_map
    139139            = new PerVolumeCatalogLangSequenceFileMap(_input_dir,_verbosity,strict_file_io);
    140         JavaRDD<String> catalog_lang_list = json_text_sample_rdd.map(volume_catalog_langfreq_map);
     140        JavaRDD<String> catalog_lang_list = json_text_rdd.map(volume_catalog_langfreq_map);
    141141        //catalog_lang_list.persist(StorageLevel.MEMORY_AND_DISK());
    142142        catalog_lang_list.setName("catalog-lang-stream");
Note: See TracChangeset for help on using the changeset viewer.