arro3-compute

0.8.0 verified Mon Apr 27 auth: no python

A compute library for the Arrow in-memory columnar format, providing high-performance operations on array and chunked arrays. Version 0.8.0 requires Python >=3.9. Development is active with no major breaking changes recently reported.

pip install arro3-compute

Common errors

error ModuleNotFoundError: No module named 'arro3_core' ↓

cause Attempting to import with underscore instead of dot: 'import arro3_core'

fix

Use correct import: import arro3.core as a3c

error AttributeError: module 'arro3.compute' has no attribute 'sum' ↓

cause The function was not found; possible typo or older version where function was in core.

fix

Upgrade to latest version: pip install --upgrade arro3-compute

error ImportError: DLL load failed while importing arro3: The specified module could not be found. ↓

cause Missing C++ runtime libraries (e.g., Visual C++ Redistributable on Windows).

fix

Install Microsoft Visual C++ Redistributable for Visual Studio 2015-2022.

Warnings

gotcha The package 'arro3-compute' is distributed separately from 'arro3-core'. If you install only 'arro3-compute', you must also install 'arro3-core' explicitly or rely on dependency resolution. ↓

fix Install both: pip install arro3-core arro3-compute

deprecated Some functions that were in arro3-core in older versions (e.g., arithmetic operations) have moved to arro3-compute. Ensure you import from the correct module. ↓

fix Use import arro3.compute for compute operations; import arro3.core for array creation.

gotcha Underlying data types (e.g., int64 vs int32) affect compute results (e.g., overflow may occur silently). ↓

fix Specify dtype explicitly when creating arrays: a3c.array([1,2,3], dtype=a3c.int64())

Imports

arro3.compute
wrong
```
import arro3_compute
```
correct
```
import arro3.compute
```
The package uses dot notation, not underscore, for import.

Quickstart

Create an array and compute its sum using arro3-compute.

import arro3.core as a3c
import arro3.compute as a3comp

arr = a3c.array([1, 2, 3, 4, 5])
result = a3comp.sum(arr)
print(result)  # 15