-
Notifications
You must be signed in to change notification settings - Fork 28
Closed
Description
What happens?
This test passes if Pandas is installed, yet fails if Pandas is not installed.
If Pandas is not installed, the result is:
[("{'3': 0}",), ("{'2': 1}",), ("{'1': 2}",)]
if Pandas is installed, the result is
[({'3': 0},), ({'2': 1},), ({'1': 2},)]
This test is from tests/fast/numpy/test_numpy_new_path.py
, which fails when Pandas is not installed.
To Reproduce
Run this code with and without Pandas.
import numpy as np
import duckdb
print(np.__version__)
print(duckdb.__version__)
duckdb_cursor = duckdb.connect()
z = []
for i in range(3):
z.append({str(3 - i): i})
z = np.array(z)
res = duckdb_cursor.sql("select * from z").fetchall()
print(res)
assert res == [
({'3': 0},),
({'2': 1},),
({'1': 2},),
]
Tested with a range of different numpy and pandas and duckdb versions, the only consistent thing with Pandas being installed or not installed.
OS:
Windows & Linux
DuckDB Package Version:
1.4.0dev (and prior versions)
Python Version:
3.13
Full Name:
Paul T
Affiliation:
Iqmo
What is the latest build you tested with? If possible, we recommend testing with the latest nightly build.
I have tested with a nightly build
Did you include all relevant data sets for reproducing the issue?
Yes
Did you include all code required to reproduce the issue?
- Yes, I have
Did you include all relevant configuration to reproduce the issue?
- Yes, I have
Metadata
Metadata
Assignees
Labels
No labels