anishmehta24 opened a new issue, #40504:
URL: https://github.com/apache/beam/issues/40504
### What happened?
With the DataFrame API, `agg` with a list that includes `'count'` or
`'size'` gives the right numbers, but those rows come back labelled `sum`
instead of `count`/`size`.
```python
import apache_beam as beam
import numpy as np
import pandas as pd
from apache_beam.dataframe.convert import to_dataframe, to_pcollection
pdf = pd.DataFrame({'x': [1.0, np.nan, 3.0], 'y': [10, 20, 30]})
print(pdf.agg(['count', 'mean']))
with beam.Pipeline() as p:
df = to_dataframe(p | beam.Create([pdf]), proxy=pdf[:0])
_ = (
to_pcollection(
df.agg(['count', 'mean']),
include_indexes=True,
yield_elements='pandas')
| beam.Map(print))
```
pandas:
```
x y
count 2.0 3.0
mean 2.0 20.0
```
Beam:
```
x y
sum 2.0 3.0
mean 2.0 20.0
```
`Series.agg(['count'])` and `Series.agg(['size'])` have the same problem.
count and size are computed per partition and then combined with a sum. When
`func` is a list, `DeferredSeries.aggregate` does that last step with
`s.agg(['sum'])`, so the row label becomes `sum`
(sdks/python/apache_beam/dataframe/frames.py, the `agg_func = ['sum'] if
isinstance(func, list) else 'sum'` line).
Seen on Beam 2.77.0 and on master (7d93ef8e76a), pandas 2.3.3, Python 3.12.
### Issue Priority
Priority: 2 (default / most bugs should be filed as P2)
### Issue Components
- [x] Component: Python SDK
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
