Closed theroggy closed 1 year ago
When reading a .gpkg file with an integer column with values > the max 32 bit int value (2.147.483.647), those values overflow to negative values when read.
The use_arrow=True code path doesn't have this issue.
use_arrow=True
Code to reproduce behaviour
from pathlib import Path import geopandas as gpd import pyogrio import shapely script_dir = Path(__file__).resolve().parent path = script_dir / "int_overflow_tmp.gpkg" gdf = gpd.GeoDataFrame( data={"overflowing_int": [2192502720]}, geometry=[shapely.Point(0,1)], crs="epsg:31370" ) pyogrio.write_dataframe(gdf, path) gdf = pyogrio.read_dataframe(path, use_arrow=False) print(f"with use_arrow=False:\n{gdf}") gdf = pyogrio.read_dataframe(path, use_arrow=True) print(f"with use_arrow=True:\n{gdf}")
When reading a .gpkg file with an integer column with values > the max 32 bit int value (2.147.483.647), those values overflow to negative values when read.
The
use_arrow=True
code path doesn't have this issue.Code to reproduce behaviour