I have written a Pig script that will perform some image processing via Python UDFs.
After doing some operations, I have something like (for example):
A = load 'data.txt' using PigStorage('|') as (name:chararray, pixelIntensity:float);
B = group A by pixelIntensity;
dump B;
B is then something like this:
(131.0,{(image1,jpg,131.0), (image2.jpg,131.0), (image3.jpg,131.0)})
(140.0,{(image5.jpg,140.0), (image5.jpg,140.0)})
(150.0,{(image4.jpg,150.0})
If I were to go
dump A;
I'd get the following:
(image1.jpg,131.0)
(image2.jpg,131.0)
(image3.jpg,131.0)
(image4.jpg,150.0)
(image5.jpg,140.0)
So I've basically grouped them using their average pixel intensity as the key.
My question is this:
Am I able to extract only 1 element from each row in B? So for example, I'll have like
(image1.jpg,131.0)
(image4.jpg,150.0)
...
A nested FOREACH
with a LIMIT
should do what you want:
A = LOAD 'data' using PigStorage(',') AS (name:chararray,pixelIntensity:float);
B = GROUP A BY pixelIntensity;
C = FOREACH B {
D = LIMIT A 1;
GENERATE flatten(D);
};
STORE C INTO 'res';