5
votes

I have a parquet table in hive with a field of type

decimal (10,2)

Now when I am reading this table in pig using parquet.pig.ParquetLoader , decimal field is getting converted to bytearray and giving junk values. Is there any solution to this issue?

1
While loading the data can you specify the field type as double or bigdecimal ? - Murali Rao
I can not change the business schema Murali and when we load a parquet file in pig, we don't specify schema. Its already present in the parquet file metadata. - Aviral Kumar
which version of hive are you using? - Frederic
I don't remember the hive version.. But its the latest one .. With cloudera 5.4 - Aviral Kumar

1 Answers

0
votes

We have written a cusotom ParquetLoader in which we extended LoadFunc class and implemented the cases we wanted for handling decimal.