Parquet troubles with decimal in Azure Data Factory V2 - azure-data-factory-2

Question

Since 3 or 4 days i'm experiencing troubles in writing decimal values in parquet file format with Azure Data Factory V2.
The repro steps are quite simple, from an SQL source containing a numeric value i map it to a parquet file using the copy activity.
At runtime the following exception is thrown:
{
"errorCode": "2200",
"message": "Failure happened on 'Source' side. ErrorCode=UserErrorParquetTypeNotSupported,'Type=Microsoft.DataTransfer.Common.Shared.HybridDeliveryException,Message=Decimal Precision or Scale information is not found in schema for column: ADDRESSLONGITUDE,Source=Microsoft.DataTransfer.Richfile.ParquetTransferPlugin,''Type=System.InvalidCastException,Message=Object cannot be cast from DBNull to other types.,Source=mscorlib,'",
"failureType": "UserError",
"target": "Copy Data"
}
In the source the complaining column is defined as numeric(32,6) type.
I think the problem is circumscribed to the parquet sink because changing the destination format to csv result in a succeeded pipeline.
Any suggestions?
Based on Jay's answer, here is the whole dataset :
SELECT
[ADDRESSLATITUDE]
FROM
[dbo].[MyTable]

Answer 1

Based on the SQL Types to Parquet Logical Types and Data type mapping for Parquet files in data factory copy activity,it supports Decimal data type.Decimal data is converted into binary data type.
Back to your error message:
Failure happened on 'Source' side.
ErrorCode=UserErrorParquetTypeNotSupported,
'Type=Microsoft.DataTransfer.Common.Shared.HybridDeliveryException,
Message=Decimal Precision or Scale information is not found in schema
for column:
ADDRESSLONGITUDE,Source=Microsoft.DataTransfer.Richfile.ParquetTransferPlugin,''
Type=System.InvalidCastException,Message=Object cannot be cast from
DBNull to other types.,Source=mscorlib,'
If your numeric data has null value, it will be converted into Int data type without any
Decimal precision or scale information.
Csv format does not have this transformation process so you could set default value for your numeric data.