Remove duplicated rows and add a partition in a hive table

remove_duplicates.hql

CREATE TABLE temp1 (anomes INT, cnpj STRING, feat0 STRING, feat1 STRING);

INSERT INTO temp1 VALUES
 (201901, '0', 'x', 'a'),
 (201901, '0', 'y', 'b'),
 (201901, '1', 'v', 'c'),
 (201901, '1', 'w', 'd'),
 (201902, '0', 'x', 'a'),
 (201902, '0', 'x', 'a'),
 (201902, '1', 'y', 'b'),
 (201902, '1', 'y', 'b');

CREATE TABLE temp2 (cnpj STRING, feat0 STRING, feat1 STRING)
PARTITIONED BY (anomes INT);

INSERT INTO temp2 PARTITION(anomes=201901)
SELECT cnpj, feat0, feat1
FROM (
  SELECT *, row_number() OVER (PARTITION BY cnpj) AS rnk
  FROM temp1
  WHERE anomes=201901
) t0
WHERE rnk=1;

INSERT INTO temp2 PARTITION(anomes=201902)
SELECT cnpj, feat0, feat1
FROM (
  SELECT *, row_number() OVER (PARTITION BY cnpj) AS rnk
  FROM temp1
  WHERE anomes=201902
) t0
WHERE rnk=1;