For some reason, I’m unable to use pandas function pandas.DataFrame.to_gbq (here).

I’ve been building this small example to reproduce the issue :

table_schema = [{'name':'company','type':'STRING'},
    {'name':'country','type':'STRING'},
    {'name':'revenue','type':'NUMERIC'},
    {'name':'pageviews','type':'INTEGER'}]
dataframe = pandas.DataFrame({
    'company': ['stackoverflow'],
    'country': ['france'],
    'revenue': [1000.23],
    'pageviews': [10000]})

table_name = 'my_dataset.petit_test'

dataframe.to_gbq(table_name,
      'PROJECT_ID',
      chunksize=100000,
      if_exists='replace',
      table_schema=table_schema)

As a result, I get a big error message ending as follow:

pandas_gbq.exceptions.ConversionError: Could not convert DataFrame to Parquet.

I can’t figure out what is going wrong here.

Does anyone get the understanding ?