Skip to content

Commit b82cd06

Browse files
authored
Rename rerun-sdk[datafusion] to rerun-sdk[dataplatform] and add pandas dependency
### Related - closes https://linear.app/rerun/issue/RR-3532/pyarrow-needs-pandas-for-nanosecond-resolution - related https://linear.app/rerun/issue/RR-3786/remove-the-pandas-dependency ### What Title. The reason for that is that pyarrow soft-depends on pandas to properly handle nanosecond timestamps (see RR-3532 for details). Renamed the extra to `dataplatform` because is just makes more sense. The `datafusion` extra is kept for backward compatibility. Source-Ref: 1cd0e1d0ab8d457c7663eeec90e3d390a99eba4a
1 parent 3f3f121 commit b82cd06

File tree

3 files changed

+23
-3
lines changed

3 files changed

+23
-3
lines changed

docs/content/overview/installing-rerun/python.md

Lines changed: 5 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -8,9 +8,13 @@ The Python SDK includes both the SDK and the Viewer, so you're ready to go with
88
- `pip install rerun-sdk` via pip
99
- `conda install -c conda-forge rerun-sdk` via Conda
1010

11+
<!-- NOLINT_START -->
12+
1113
Conda always comes with support for all features but if using pip you may need to specify optional features:
1214
- `pip install rerun-sdk[notebook]` for the embedded notebook tools
13-
- `pip install rerun-sdk[datafusion]` for the query api tools
15+
- `pip install rerun-sdk[dataplatform]` for the query api tools
16+
17+
<!-- NOLINT_END -->
1418

1519
## Next steps
1620

rerun_py/pyproject.toml

Lines changed: 7 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -38,6 +38,7 @@ text = "MIT OR Apache-2.0"
3838
tests = [
3939
"inline-snapshot==0.31.1",
4040
"opencv-python>4.6",
41+
"pandas>=2",
4142
"polars==1.36.1",
4243
"pytest==8.4.2",
4344
"semver>=3.0,<3.1",
@@ -47,7 +48,12 @@ tests = [
4748
"datafusion==50.1.0",
4849
]
4950
notebook = ["rerun-notebook==0.30.0-alpha.1+dev"]
50-
datafusion = ["datafusion==50.1.0"]
51+
# TODO(RR-3786): when pyarrow is fixed, we should remove the pandas dependency
52+
datafusion = ["datafusion==50.1.0", "pandas>=2"] # deprecated, replaced by `dataplatform` (NOLINT)
53+
dataplatform = [ # NOLINT
54+
"datafusion==50.1.0",
55+
"pandas>=2",
56+
] # pandas is needed so pyarrow properly converts nanoseconds (RR-3532)
5157
# Note: We avoid self-referential extras like "rerun-sdk[notebook]" because they cause
5258
# uv to fetch rerun-sdk from PyPI when the local package isn't built yet.
5359
all = ["rerun-notebook==0.28.0-alpha.1+dev", "datafusion==50.1.0"]

uv.lock

Lines changed: 11 additions & 1 deletion
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.

0 commit comments

Comments
 (0)