lib-polarslisted

The polars wheel is now an empty 865 KB py3-none-any shim hard-pinned to polars-runtime-32, so a lockfile listing only polars does not pin the engine. TRIGGER - polars, "import polars as pl", LazyFrame, scan_parquet, collect(), pl.col, join_asof, group_by, with_columns, polars-runtime-32, polars-runtime-64, polars-lts-cpu, polars 2.0.0rc1, "pip download polars", vendored or air-gapped polars install, polars wheel has no compiled code, porting pandas merge_asof to polars.join_asof, polars sortedness. The runtime split landed at 1.34.0b2 on 2025-09-26, so install matrices, wheel audits and lockfiles written from memory are wrong. SKIP for market-data-engineering, the skill for storage formats and time-series stores. SKIP when the question is WHICH library to choose, or names no library at all - both belong to the domain skill.
howard-lynn-ye/fin-skills · ★ 1 · AI & Automation · score 77

Install: claude install-skill howard-lynn-ye/fin-skills

# polars The best single-node dataframe default for large panels — multi-threaded Rust, Arrow columnar, a real lazy optimiser, and it releases the GIL. It also changed its own packaging, and its as-of join is now the unsafe one. | | | |---|---| | pip / import | `polars` / `import polars as pl` — **plus `polars-runtime-32`, which is not optional** | | Version | 1.44.1 (2026-08-26) · 432 releases · **2.0.0rc1 shipped 2026-09-02** | | Licence | MIT (both `polars` and `polars-runtime-32`) | | Python | `>=3.10`. `polars` ships **1 wheel, `py3-none-any`**; `polars-runtime-32` ships 9 `cp310-abi3` wheels | | Status | 39,646★, very active. **Pin `polars<2`** until you have read the 2.0 migration guide | ## The trap that costs you money **`polars-1.44.1-py3-none-any.whl` is 865,208 bytes and contains no compiled code.** Its `Requires-Dist` opens with an unconditional hard pin: ``` polars-runtime-32==1.44.1 # mandatory, NOT an extra polars-runtime-64==1.44.1 ; extra == "rt64" polars-runtime-compat==1.44.1 ; extra == "rtcompat" ``` The engine lives in `polars-runtime-32` — MIT, `cp310-abi3` wheels (one covers Python 3.10 through 3.14+) for macOS x86_64/arm64, manylinux + musllinux x86_64/aarch64, win_amd64/arm64, 47–54 MB each. The split landed at **1.34.0b2 (2025-09-26)**, a 5,966-byte wheel where 1.34.0b1 three days earlier still shipped ~40 MB binaries; everything from 1.34 on is split. Corroborating: `polars-lts-cpu` is frozen at **1.33.1 (2025-09-09)**,