Yikun commented on code in PR #36793:
URL: https://github.com/apache/spark/pull/36793#discussion_r894971731
##########
python/pyspark/sql/session.py:
##########
@@ -952,12 +953,29 @@ def createDataFrame( # type: ignore[misc]
schema = [x.encode("utf-8") if not isinstance(x, str) else x for x
in schema]
try:
- import pandas
+ import pandas as pd
has_pandas = True
except Exception:
has_pandas = False
- if has_pandas and isinstance(data, pandas.DataFrame):
+
+ try:
+ import numpy as np
+
+ has_numpy = True
+ except Exception:
+ has_numpy = False
+
+ if has_numpy and isinstance(data, np.ndarray):
+ from pyspark.sql.pandas.utils import require_minimum_pandas_version
+
+ require_minimum_pandas_version()
+ if data.ndim not in [1, 2]:
+ raise ValueError("NumPy array input should be of 1 or 2
dimensions.")
+ column_names = ["value"] if data.ndim == 1 else ["_1", "_2"]
+ data = pd.DataFrame(data, columns=column_names)
Review Comment:
question: Is other numpy types supported in future also considered first
covert to pandas frame first? numpy --> pandas --> spark df?
##########
python/pyspark/sql/session.py:
##########
@@ -952,12 +953,29 @@ def createDataFrame( # type: ignore[misc]
schema = [x.encode("utf-8") if not isinstance(x, str) else x for x
in schema]
try:
- import pandas
+ import pandas as pd
has_pandas = True
except Exception:
has_pandas = False
- if has_pandas and isinstance(data, pandas.DataFrame):
+
+ try:
+ import numpy as np
+
+ has_numpy = True
+ except Exception:
+ has_numpy = False
+
+ if has_numpy and isinstance(data, np.ndarray):
+ from pyspark.sql.pandas.utils import require_minimum_pandas_version
+
+ require_minimum_pandas_version()
Review Comment:
nit: If only numpy installed but pandas not installed, will only raised
pandas not installed. User may confused why need to install pandas but I just
want to using numpy ?
So maybe give a certainly exceptions before here to tell users, numpy type
will first convert pandas df in pyspark so pandas installed is required, like:
```python
if not has_pandas:
// warning or raised friendly exception
from pyspark.sql.pandas.utils import require_minimum_pandas_version
require_minimum_pandas_version()
```
or add a notes before here at least.
##########
python/pyspark/sql/session.py:
##########
@@ -952,12 +953,29 @@ def createDataFrame( # type: ignore[misc]
schema = [x.encode("utf-8") if not isinstance(x, str) else x for x
in schema]
try:
- import pandas
+ import pandas as pd
has_pandas = True
except Exception:
has_pandas = False
- if has_pandas and isinstance(data, pandas.DataFrame):
+
+ try:
+ import numpy as np
Review Comment:
Should we add `numpy` as requirement in
[setup.py](https://github.com/apache/spark/blob/master/python/setup.py#L266-L269)?
and mentioned it in deps doc?
[1] https://github.com/apache/spark/blob/master/python/setup.py#L266-L269
[2]
https://dist.apache.org/repos/dist/dev/spark/v3.3.0-rc1-docs/_site/api/python/getting_started/install.html#dependencies
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
