Pig : UDFs

Pig UDFS ----------   UDF ---> user defined functions.      adv:        i)  custom functionalities.       ii)  reusability.  Pig UDFs can be developed by     java    python     ruby     c++     javascript     perl step1:    Develop udf code. step2:    export into jar file    ex: /home/cloudera/Desktop/pigs.jar step3:    register jar file into pig.  grunt> register Desktop/pigs.jar step4:    create temporory function  for udf class.  grunt> define  ucase pig.analytics.ConvertUpper(); step5:   calling the function:  grunt>e =  foreach emp generate       id, ucase(name) as name, sal,         ucase(sex) as sex, dno;   package  pig.analytics; import ..... --> ucase(name) ---> upper conversion public class ConvertUpper extends EvalFunc   {      public String exec(Tuple v)       throws IOException      {         String str = (String)v.get(0);         String res = str.toUpperCase();         retrun res;                  }    } -------------------------- $ cat > samp 100,230,400 123,100,90 140,560,430 $ hadoop fs -copyFromLocal samp piglab grunt> s = load 'piglab/samp'             using PigStorage(',')           as (a:int, b:int, c:int); package pig.analytics; .... public class RMax extends EvalFunc {     public  Integer exec(Tuple v)      throws IOException     {       int a =(Integer) v.get(0);       int b =(Integer) v.get(1);       int c =(Integer) v.get(2);       int big = a; // 10,20,3       if (a>big) big = a;       if (b>big) big = b;       if (c>big) big = c;       return  big;     }  } export into jar : Desktop/pigs.jar grunt> register Desktop/pigs.jar; grunt> define rmax pig.analytics.RMax(); grunt> res = foreach s generate *,                    rmax(*) as max; --------------------------------  package pig.analytics;  .......  public class RowMax    extends EvalFunc  {     public Integer exec(Tuple v) throws IOException     {      List lobs = v.getAll() ;      int max = 0;      int cnt =0;     // -20,-3,-40      for(Object o : lobs)      {        cnt++;        int val = (Integer)o;        if(cnt==1) max = val;        max = Math.max(max, val);      }      return max;     }  } export in to jar : Desktop/pigs.jar grunt> register Desktop/pigs.jar grunt> define dynmax pig.analytics.RowMax(); grunt> r = foreach s generate *, dynmax(*) as m; ----------------------------------------- emp = load 'piglab/emp' using PigStorage(',')    as (id:int, name:chararray, sal:int,      sex:chararray, dno:int); grade() dname() gender() package pig.analytics; public class Gender extends EvalFunc {  public String exec(Tuple v) throws IOException  {      String s =(String) v.get(0);      s = s.toUpperCase();      if (s.matches("F"))        s = "Female";      else        s = "Male";      return s;  } } ----------------- package pig.analytics; public class Grade extends EvalFunc {  public String exec(Tuple v) throws IOException  {      String sal =(Integer) v.get(0);      String grade;      if (sal>=70000)        grade="A";      else if (sal>=50000)           grade="B";          else if (sal>=30000)                grade="C";               else                grade="D";      return grade;  } } ------ package pig.analytics; public class DeptName extends EvalFunc {  public String exec(Tuple v) throws IOException  {     int dno = (Integer)v.get(0);     String dname;     switch (dno){     case 11 :           dname = "Marketing";           break;     case 12 :           dname = "HR";           break;     case 13 :           dname = "Finance";           break;     default:           dname = "Others";      }     return dname;    } } --------------------------------- --------------------------- export into jar : Desktop/pigs.jar; grunt> register Desktop/pigs.jar; grunt> define gender pig.analytics.Gender(); grunt> define grade pig.analytics.Grade(); grunt> define dept pig.analytics.DeptName(); grunt> res = foreach emp generate     id, ucase(name) as name,      sal, grade(sal) as grade,     gender(sex) as sex,     dept(dno) as dname ; ---------------------------------