va-framework

KO | EN

gitlite — search

#react #ai #llm #golang #go #machine-learning #docker #macos #kubernetes #vue #awesome-list #pytorch #openai #open-source

(★ 26)

A framework designed for jump-starting a visual analytics system.

va-framework 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# va-framework v2

`main` now tracks the React-native rewrite of the visual analytics framework.
The stable Vue release is preserved on the `release/v1.0.0` branch.

## Current milestone

`v2.3.18` is the gallery hardening and contribution-architecture release.

The home route at `/` remains the graph workbench. This final `v2.3` release closes the gallery line with provenance, contributor guidance, and smoke coverage around the example registry:

- every gallery-backed example shell now renders a shared provenance panel that shows both example-source and dataset-source attribution
- the gallery line now includes contributor-facing guidance for adding datasets, visualization adapters, and shared controls in `docs/gallery-contributors.md`
- route and registry smoke tests now cover the main gallery surfaces and ensure every catalog example resolves to a registered dataset
- the `stocks` focus-context example from `v2.3.17` remains part of the final gallery line alongside graph, hierarchy, SPLOM, and Sankey examples

The `v2.3` line now consists of:

- `v2.3.0`: graph-native query contracts, local graphology runtime, remote graph queries, and the first graph workspace
- `v2.3.1`: graph shell polish, full-graph default scope, visible route switching, and drawer cleanup
- `v2.3.2`: graph workbench foundation, URL-backed technique switching, and hierarchy dataset support
- `v2.3.3`: adjacency matrix brushing plus first integrated tree and multivariate technique renderers
- `v2.3.4`: matrix ordering and graph-analysis polish
- `v2.3.5`: tree data foundation, hierarchy pathing, and flare metadata checks
- `v2.3.6`: explicit and implicit tree-technique polish with selected-path highlighting
- `v2.3.7`: multivariate field profiles, missing-value handling, and encoding groundwork
- `v2.3.8`: multivariate layout polish, legends, and facet-aware analytical summaries
- `v2.3.9`: graph workbench hardening, guidance, accessibility labels, and regression coverage
- `v2.3.10`: canonical Les Miserables dataset import, duplicate-key fix, and remote runtime cleanup
- `v2.3.11`: example gallery foundation, visualization catalog, and generic example routes
- `v2.3.12`: shared control system, command-based example switching, and gallery category tabs
- `v2.3.13`: dataset-pack import, gallery dataset cards, and placeholder example registration
- `v2.3.14`: brushable scatterplot matrix on penguins
- `v2.3.15`: native Sankey flow example on energy
- `v2.3.16`: hierarchy technique suite on flare
- `v2.3.17`: focus-plus-context time series on stocks
- `v2.3.18`: gallery hardening, provenance panels, contributor docs, and route/catalog smoke coverage

The core stack now includes:

- `TanStack Query` for async server-state work
- `Zustand` for persisted coordination state
- `DuckDB-Wasm + Apache Arrow` for tabular local analytics
- `graphology` for local graph data modeling and metrics
- custom D3 visualization primitives in `packages/vis-core`

The release ladder remains:

- `v2.0.0`: React shell and monorepo baseline
- `v2.1.0`: data foundation with shared contracts, coordinated state, and local/remote execution
- `v2.2.0`: single-view analytics
- `v2.3.0`: graph data
- `v2.3.1`: graph shell polish
- `v2.3.2`: graph workbench foundation
- `v2.3.3`: adjacency matrix brush MVP
- `v2.3.4`: matrix ordering and graph-analysis polish
- `v2.3.5`: tree data foundation
- `v2.3.6`: tree techniques
- `v2.3.7`: multivariate network foundation
- `v2.3.8`: multivariate network techniques
- `v2.3.9`: graph workbench hardening
- `v2.3.10`: canonical Les Miserables dataset and remote runtime fixes
- `v2.3.11`: example gallery foundation
- `v2.3.12`: shared control system
- `v2.3.13`: dataset pack import
- `v2.3.14`: brushable scatterplot matrix
- `v2.3.15`: flow example
- `v2.3.16`: tree and hierarchy suite
- `v2.3.17`: time-series example and graph polish
- `v2.3.18`: hardening and contribution architecture
- `v2.4.0`: spatio-temporal data
- `v2.5.0`: multi-view coordination
- `v2.6.0`: spatial-ready foundations

## Workspace layout

- `apps/web`: Next.js App Router frontend
- `apps/api`: FastAPI backend
- `packages/ui`: shadcn-style component baseline
- `packages/contracts`: shared TypeScript contracts and schemas
- `packages/view-system`: workspace and view contracts
- `packages/vis-core`: visualization primitives

## Development

This repo now uses `pnpm` workspaces for the JavaScript side and Python 3.11+ for the API.

```bash
pnpm install
pnpm dev:web
pnpm dev:api
```

Or run both app processes together:

```bash
pnpm dev
```

If the repo stays on a NAS or another slow filesystem, prefer the default webpack-based `pnpm dev:web`.
If you move the repo to a local SSD and want the faster compiler path, use:

```bash
pnpm dev:web:turbo
```

The frontend runs at <http://localhost:3000>.
The backend health endpoint is available at <http://127.0.0.1:8000/api/health>.

Routes:

- `/`: graph workbench
- `/gallery`: visualization gallery
- `/examples/[exampleId]`: registry-backed example route
- `/cars`: cars single-view reference workflow

The graph workbench defaults to `miserables` + `force`, and also supports `flare` as the hierarchy dataset for the tree line.
The canonical Les Miserables dataset used here is the Vega sample graph with 77 nodes and 254 links.

Contributor guide:

- `docs/gallery-contributors.md`: add a dataset, add a visualization adapter, and wire shared controls into the workbench shell

## Graph workbench endpoints

- `GET /api/health`
- `GET /api/datasets`
- `POST /api/query`
- `POST /api/jobs`
- `GET /api/jobs/{id}`

va-framework

va-framework 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

va-framework

va-framework 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README