vatsan

DROP FUNCTION IF EXISTS posdemo.tag_pos(varchar);
CREATE FUNCTION posdemo.tag_pos(varchar)
RETURNS SETOF token_tag
AS
'postagger.nlp.POSTagger.tagTweet'
IMMUTABLE LANGUAGE PLJAVAU;

vatsan

 
 package postagger.util;
 /**
 * A class to hold the {token, index, tag} triplet.
 * @author Srivatsan Ramanujam<vatsan.cs@utexas.edu>
 *
 */
 public class TaggedResult {
      private int index;
      private String token;

vatsan

package postagger.util;
import java.sql.ResultSet;
import java.sql.SQLException;
import java.util.ArrayList;
import java.util.Iterator;
import java.util.List;
import org.postgresql.pljava.ResultSetProvider;
 	 
/**
* Class implement PL/Java’s ResultSetProvider interface, to  handle the return of rows of 

vatsan

package postagger.nlp;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import org.postgresql.pljava.ResultSetProvider;
import postagger.util.TaggedResult;
import postagger.util.TaggedResultProvider;
//CMU Ark Tweet NLP package (GPL v2 license)
import cmu.arktweetnlp.*;
import cmu.arktweetnlp.Tagger.TaggedToken;

vatsan

select id,
       (t).indx,
       (t).token,
       (t).tag
from
(
	select id,
	posdemo.tag_pos(tweet_body) as t
	from posdemo.training_data
) q;

vatsan

id     | indx |  token   | tag
------------+------+----------+-----
1467810672 |    0 | is       | V
1467810672 |    1 | upset    | A
1467810672 |    2 | that     | P
1467810672 |    3 | he       | O
1467810672 |    4 | can't    | V
1467810672 |    5 | update   | V
1467810672 |    6 | his      | D
1467810672 |    7 | Facebook | ^

vatsan

Table "posdemo.training_data"
Column   |            Type             | Modifiers | Storage  | Description
------------+-----------------------------+-----------+----------+-------------
id         | bigint                      |           | plain    |
ts         | timestamp without time zone |           | plain    |
poster     | text                        |           | extended |
tweet_body | text                        |           | extended |
Has OIDs: no
Distributed by: (id)

vatsan

vatsandb=# select id, tweet_body from posdemo.training_data limit 5;

id     |                                                                tweet_body
------------+-----------------------------------------------------------------------------------------

1467820906 | @localtweeps Wow, tons of replies from you, may have to unfollow so I can see my friends' tweets, you're scrolling the feed a lot.
1467862806 | @MySteezRadio I'm goin' to follow u, since u didn't  LOL  GO ANGELS!
1467891880 | Argh! I was suuuper sleepy an hour ago, now I'm wide awake.  Hope I don't stay up all night. :-/
1467896211 | michigan state you make me sad 
1467911846 | @bananaface IM SORRY I GOT YOU SICK.  lol. going to bed too. NIGHT!

vatsan

public static class ImageToSequenceMapper extends Mapper<LongWritable,Text,Text, Text> {
    // The input to each mapper is a set of lines from a text file on HDFS, where each line contains a path to an HDFS image file
    // The job input format is NLineInputFormat and the job output format is TextOutputFormat
    @Override
    public void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
        // The input argument value contains the HDFS path to the image file  
        String imgName = value.toString();
        FileSystem fs = FileSystem.get(context.getConfiguration());
        FSDataInputStream in = null;

vatsan

g++ -shared -Wl,-soname,canny_edge_detection -fPIC -ocanny_edge_detection.so -lopencv_core -lopencv_imgproc -lopencv_highgui CannyEdgeDetectionCtypes.cpp