[GitHub] [arrow] AlenkaF commented on issue #12458: [Python]read_parquet throw ArrowNotImplementedError if empty dataframe exists

[GitBox <gi...@apache.org>]

AlenkaF commented on issue #12458:
URL: https://github.com/apache/arrow/issues/12458#issuecomment-1044500102


   You could also try to pass a schema to `to_parquet` if you know you have strings for certain columns:
   
   ```python
   import pandas as pd
   import pyarrow as pa
   
   dr = pd.date_range(start='2022-01-01', end='2022-01-10', freq='D')
   df = pd.DataFrame(index=dr)
   df['A'] = pd.to_datetime('today')
   df['B'] = 1.0
   df['C'] = 'a'
   df['D'] = 2
   
   schema = pa.schema([
       pa.field('A', pa.timestamp('ns')),
       pa.field('B', pa.float64()),
       pa.field('C', pa.string()),
       pa.field('D', pa.int64())])
   
   df.head(0).to_parquet('b/0.parquet', schema=schema)
   df.to_parquet('b/1.parquet')
   pd.read_parquet('b') 
                              A    B  C  D
   0 2022-02-18 14:03:36.691288  1.0  a  2
   1 2022-02-18 14:03:36.691288  1.0  a  2
   2 2022-02-18 14:03:36.691288  1.0  a  2
   3 2022-02-18 14:03:36.691288  1.0  a  2
   4 2022-02-18 14:03:36.691288  1.0  a  2
   5 2022-02-18 14:03:36.691288  1.0  a  2
   6 2022-02-18 14:03:36.691288  1.0  a  2
   7 2022-02-18 14:03:36.691288  1.0  a  2
   8 2022-02-18 14:03:36.691288  1.0  a  2
   9 2022-02-18 14:03:36.691288  1.0  a  2
   ```
   
   See https://issues.apache.org/jira/browse/ARROW-14488


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org