Purpose: Sometime you prefer to access your data, which resides in BigQuery locally using pandas.
Usage:
bqd = bqdask.BQDask(project_id='Google Cloud Project', path_parquet='Path where table date should be saved', export_bucket='Goocle Cloud Storage Bucket name', service_acct_json='Service account.json path', bq_location='EU')
def df_fixer(df):
df = df.replace({'somefield': {1.0: True, 0.0: False}})
#required for incremental sync to work.
df.increment_stamp = dd.to_datetime(df.created_timestamp,
unit='ms')
#create partition column for partitioning
df = partition_by(df, "created_timestamp")
return df
refresh_table(table='bq_project.bq_dataset.bq_table', query_increment_column='created_timestamp',
unique_column='id',
limit_columns='id,field1,field2,field3'.split(
','), df_fixer=df_fixer)