# PyArrow

6 Hacker News commenters recommended PyArrow and 0 criticized it, November 2025 to November 2026. It was recommended in 4 of 13 months.

**Verdict:** People recommend PyArrow as a strong modern choice for working with data, often alongside Polars and DuckDB, and for its better data type handling. The most common caveat is that it is very heavy.

## Why people recommend it

- Pairs well with Polars and DuckDB (2 comments): "Everything that I build greenfield moving forward I plan to use DuckDB, Polars, or PyArrow." ([source](https://news.ycombinator.com/item?id=49981111)) "Polars+pyarrow+ibis have replaced most of my pandas usage." ([source](https://news.ycombinator.com/item?id=46355085))
- Seen as a good modern data choice (2 comments): "Everything that I build greenfield moving forward I plan to use DuckDB, Polars, or PyArrow." ([source](https://news.ycombinator.com/item?id=49981111)) "I totally agree that these are the good choices" ([source](https://news.ycombinator.com/item?id=49982039))

## What people warn about

- Very heavy package (1 comments): "Pyarrow is great but it is very heavy." ([source](https://news.ycombinator.com/item?id=49985949))
- Best with little data transformation (1 comments): "straight up arrow if you aren't doing much transformation of the data" ([source](https://news.ycombinator.com/item?id=49982350))

## Alternatives hackers recommend

- [Polars](https://hackerapproved.com/p/polars.md): 152 recommend
- [pandas](https://hackerapproved.com/p/pandas.md): 93 recommend
- [NumPy](https://hackerapproved.com/p/numpy.md): 86 recommend
- [SciPy](https://hackerapproved.com/p/scipy.md): 12 recommend
- [sh Python library](https://hackerapproved.com/p/sh.md): 10 recommend

## Cite as

Hacker Approved, "PyArrow", https://hackerapproved.com/p/pyarrow/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).