Apache pig 连接来自不同关系的字段

有两种关系

r1: {f1: chararray, f2: chararray}
r2: {f3: chararray, f4: chararray}

这两个关系中都没有唯一的键，但元组的数量相同

---

是否有一种方法可以连接关系的相应字段以获得类似f2、f4的输出？

如果元组顺序正确，可以使用秩

r1a = RANK r1 BY * DENSE;
r2a = RANK r2 BY * DENSE;

r1r2 = JOIN r1a BY $0, r2a BY $0;

---

另一种选择是使用：交叉（）

注意：从文档中提取：CROSS是一项昂贵的操作，应谨慎使用

清管器脚本：

R1 = LOAD 'a.csv'  USING  PigStorage(',') AS (f1:chararray,f2:chararray);
R2 = LOAD 'b.csv'  USING  PigStorage(',') AS (f3:chararray,f4:chararray);

R3 = CROSS R1,R2;

R4 = FOREACH R3 GENERATE f2,f4;

DUMP R4;

f1_value,f2_value

(f2_value,f4_value)

输入：

R1 = LOAD 'a.csv'  USING  PigStorage(',') AS (f1:chararray,f2:chararray);
R2 = LOAD 'b.csv'  USING  PigStorage(',') AS (f3:chararray,f4:chararray);

R3 = CROSS R1,R2;

R4 = FOREACH R3 GENERATE f2,f4;

DUMP R4;

f1_value,f2_value

(f2_value,f4_value)

a.csv:

R1 = LOAD 'a.csv'  USING  PigStorage(',') AS (f1:chararray,f2:chararray);
R2 = LOAD 'b.csv'  USING  PigStorage(',') AS (f3:chararray,f4:chararray);

R3 = CROSS R1,R2;

R4 = FOREACH R3 GENERATE f2,f4;

DUMP R4;

f1_value,f2_value

(f2_value,f4_value)

b.csv

f3_Value,f4_value

输出：转储R4:

R1 = LOAD 'a.csv'  USING  PigStorage(',') AS (f1:chararray,f2:chararray);
R2 = LOAD 'b.csv'  USING  PigStorage(',') AS (f3:chararray,f4:chararray);

R3 = CROSS R1,R2;

R4 = FOREACH R3 GENERATE f2,f4;

DUMP R4;

f1_value,f2_value

(f2_value,f4_value)

---

连接不应该是$0吗？交叉不起作用，因为您不希望完全交叉连接，而只希望连接具有相同行号的行。