/jump to repo

Repos Recent Bookmarks Watched Notes Tags Discover Compare Stats GitHub

Repository brief

pandas-dev/pandas

Read the upstream summary on the left, browse the cached forks below it, and load each fork comparison into the right-hand panel.

Cached analysis

cached 2026-03-30T11:46:32.904Z

3mo ago

pandas-dev/pandas

pandas-dev/pandas is the main pandas repository: a stable, actively maintained Python data analysis/manipulation library for labeled tabular data, time series, and statistics. It is highly popular and well-established, with 48,284 stars, 19,795 forks, and recent commits on 2026-03-29 and 2026-03-30.

Loading tags...

Stars48,284

Forks19,795

Default branchmain

Last pushed2026-03-30T11:22:48Z

Recommended shortcuts

Jump straight into Discofork's strongest cached fork picks, or open a compare view in one click.

Forks

Choose a fork to inspect

10 of 10 fork briefs

Maintenance:

Magnitude:

Sort:

Selected

Prefer upstream pandas unless you specifically need an old, divergent codebase for legacy compatibility or historical reasons. This fork is not a good choice for users who want current pandas features, active maintenance, or low-risk adoption.

Prefer upstream pandas unless you specifically need this fork's custom, legacy behavior. Choose this fork only if its local changes are required and you are prepared to own the maintenance burden.

Choose this fork only if you specifically need its older 2019-era behavior or one of its backported fixes. For most adopters, upstream pandas is the better choice because this fork is heavily stale and far behind current maintenance.

Prefer upstream pandas for new work. Choose this fork only if you need its legacy behavior or specific fork-local changes and are prepared to maintain a large, stale divergence yourself.

Prefer this fork only if you need its custom local changes and are willing to carry a large divergence from upstream. For most users, upstream pandas is the safer choice because it is far more current and better maintained.

Prefer this fork only if you are locked to legacy pandas behavior; otherwise upstream pandas is the better choice because this fork is heavily out of date and materially diverged.

Prefer this fork only if you specifically want its LLM-processing/examples angle and can tolerate major loss of upstream pandas completeness. For general pandas use, upstream is the better choice.

Prefer upstream pandas unless you explicitly need this fork's legacy behavior or custom internals; the fork is materially outdated and highly divergent, so adoption is only sensible for compatibility-constrained projects that can accept ongoing maintenance work.

Choose this fork only if you want a customized pandas base and are prepared to own divergence. For most adopters who want current, stable pandas behavior with minimal maintenance overhead, upstream is the safer choice.

Prefer upstream pandas unless you specifically need legacy behavior or this fork's older demo/compatibility changes. For most adopters, the fork's staleness and distance from current pandas outweigh its narrow advantages.

Fork comparison

wesm/pandas

38/100

stale

significant_divergence

Prefer upstream pandas unless you specifically need an old, divergent codebase for legacy compatibility or historical reasons. This fork is not a good choice for users who want current pandas features, active maintenance, or low-risk adoption.

Likely purpose

A historical or experimental pandas line for users who want an older, customized base rather than the current upstream release stream.

Best for

Teams maintaining legacy software pinned to an older pandas behavior set.; Developers investigating historical pandas internals or experimenting with alternate core behaviors.; Users who explicitly need a fork with long-frozen dependencies and do not want upstream churn.

Additional features

Custom forked behavior across core DataFrame, Series, Index, groupby, time series offset, SQL, PyTables, and styling code paths.
Substantial test and documentation divergence, suggesting the fork was adapted for a specific downstream workflow or compatibility goal.
A separate dependency/environment snapshot (`pixi.lock`) and altered build/runtime context, which points to a curated frozen environment rather than tracking upstream.
Expands API, schema, or SDK surface area for downstream integrations and generated clients.
Changes agent workflow behavior, such as prompt handling, handoffs, worktrees, memory, or orchestration flow.

Missing features

It is missing years of upstream pandas development, including current bug fixes, deprecations, and API refinements from 2020-2026.
Modern upstream build and packaging expectations are likely absent or outdated relative to the current `main` branch.
Current upstream ecosystem support and maintenance around filesystems, cloud storage, databases, parquet/feather, and newer optional dependency groups is unlikely to be fully present.
Recent upstream fixes in datetime, timedelta, SQL, Stata, and indexing behavior are not reflected here.
Users will not get the current level of active CI, testing, and documentation maintenance seen in upstream.
Some upstream behavior appears to be removed, disabled, or intentionally stripped down in this fork.

Strengths

May preserve older APIs or behaviors that some legacy workflows still depend on.
Could be attractive for teams that need a frozen historical baseline or want to study earlier pandas implementation choices.
The fork is not archived, so it may still be intentionally maintained for a narrow purpose even if it is stale relative to upstream.

Risks

Large code divergence makes upstream merges and security/bugfix backports difficult.
The fork is many years behind upstream, so compatibility with modern Python/package stacks is a real risk.
Heavy deletions suggest behavior removal or simplification, which can break existing pandas workflows.
Because the last push is from 2019, adopters should assume substantial unreviewed drift from current pandas semantics.