Unable to create partitioned tables...again

[wenhsinhu]

Hi,

This issue is related to #359.

I tried to generate a partitioned table with the solutions within the above issue and encountered some problems. Not sure if this is caused by version updates.

1. Specifiy the fields as json: for a json like this:

name = c('date','dummy_columnA','dummy_columnB')
type = c('DATE','STRING','INT64')
schema = jsonlite::toJSON(as.data.frame(cbind(name,type)))
schema
#> [{"name":"date","type":"DATE"},{"name":"dummy_columnA","type":"STRING"},{"name":"dummy_columnB","type":"INT64"}]

I tried the following scripts:

bigrquery::bq_table_create(x = bq_table, fields = schema,
            timePartitioning = list(timepartitioning = list(type='DAY', field= 'date'))
            )

an error returned

#> Error in UseMethod("as_bq_fields"): no applicable method for 'as_bq_fields' applied to an object of class "json"

2. Specify the field as bq_fields

bq_fields <- bigrquery::as_bq_fields(list(bigrquery::bq_field('date','DATE'),
                                          bigrquery::bq_field('dummy_columnA','STRING'),
                                          bigrquery::bq_field('dummy_columnB','INT64')))


bigrquery::bq_table_create(x = bq_table, fields = bq_fields,
                timePartitioning = list(timepartitioning = list(type='DAY', field= 'date'))
)

A partitioned table is generated, whereas the field setting is not working.

NOTES: The following bq commend works for me:

bq mk \
--schema=date:DATE,dummy_colA:STRING,dummy_colB:INTEGER \
 --time_partitioning_field=date \
 --time_partitioning_type=DAY bi_report.dummy_user_profile

[byapparov]

You have double layer of timepartitoning lists. Can you double check the example from the previous issue

[hadley]

Closing this since it's been 2 years without further communication.