PL/SQL :: Fetch DISTINCT Values From Partitioned Table

What is the fastest way to fetch DISTINCT values from partitioned table?

1) DISTINCT/UNIQUE
2) GROUP BY
3) PARTITION BY OVER()
4) MAX(ROWID)

Table Definition
CREATE TABLE STG_SOS_SALES_FACT_STUDY
(
  CNTRY_KEY     NUMBER,
  STUDY_ID      NUMBER,
  PRD_KEY_YEAR  NUMBER,
  PRD_KEY_WEEK  NUMBER,
  DATE_FROM     DATE,

[Code]...

-> PARTITION BY RANGE (PRD_KEY_YEAR, PRD_KEY_WEEK)
-> SUBPARTITION BY LIST (CNTRY_KEY)

** Local Partition Indexes
1) CN_SD_CTG_PRD_PRDC_IDX = STG_SOS_SALES_FACT_STUDY (PRD_KEY_YEAR, PRD_KEY_WEEK, CNTRY_KEY, STUDY_ID, CTG_ID, PRDC_KEY)
2) CN_SD_PRD_STR_CTG_IDX = STG_SOS_SALES_FACT_STUDY (PRD_KEY_YEAR, PRD_KEY_WEEK, CNTRY_KEY, STUDY_ID, STR_KEY)#Query:
SELECT DISTINCT PRD_KEY_WEEK, PRD_KEY_YEAR

[Code]...

** Explain Plan:
Plan
SELECT STATEMENT  ALL_ROWSCost: 6,235  Bytes: 629  Cardinality: 37                           
      8 HASH UNIQUE  Cost: 6,235  Bytes: 629  Cardinality: 37                      
         7 CONCATENATION                 
              3 PARTITION RANGE ITERATOR  Cost: 1,985  Bytes: 1,031,900  Cardinality: 60,700  Partition #: 3  Partitions accessed #194 - #207          

[Code]...

Partition #: 7  Partitions determined by Key ValuesThe above query is taking around 6-7 minutes to fetch the data.