observablehq/plot
Plot
这是一个图形语法可视化库，用于通过将表格数据映射到视觉标记来构建图表。它作为一个 SVG 数据可视化工具和探索性数据分析 API，允许用户渲染复杂的可视化效果和地理地图。
该库具有一个 GeoJSON 地图渲染器，可将球坐标投影到二维像素空间，以及一个用于高效数据处理的 Apache Arrow 可视化接口。
其功能面涵盖通过分箱（binning）和分组进行数据转换、通过自动比例推断和配色方案应用进行视觉编码，以及生成小多重图（small multiples）。它支持在分层视图中渲染几何形状，并在服务器端环境中导出静态图像。
Features
- Layered Visualization Composition - Implements a layered grammar of graphics to compose independent data, mark, and scale layers into a coherent view.
- Declarative Visualization Grammars - Implements a formal grammar of graphics that builds charts by layering geometric shapes and mapping data to visual channels.
- Apache Arrow Processing - Processes diverse input structures, including high-efficiency Apache Arrow tables, for optimized data visualization.
- Data Visualization - Transforms rows of tabular data into visual representations to accelerate exploratory data analysis.
- Visual Channel Bindings - Binds data columns and accessor functions to visual mark options to vary attributes across different data points.
- Exploratory Data Analysis - Provides an API for rapidly transforming tabular data into charts to discover patterns and statistical insights.
- Exploratory Analysis APIs - Offers a specialized API for rapid data exploration through automated scales, transforms, and small multiples.
- Scale and Guide Inference Engines - Automatically infers appropriate quantitative, ordinal, and binned scales from data to generate axes and legends.
- Tabular Data Analysis - Processes tabular data through binning, grouping, and stacking to prepare it for visual representation.
- In-Memory Format Support - Handles diverse data structures, including arrays of objects and Apache Arrow tables, to improve processing efficiency.
- Tabular Data Transformations - Computes binned, grouped, or stacked values during the rendering pipeline to prepare raw data for visual representation.
- Geometric Data Mappings - Renders data as a variety of geometric marks, such as dots, lines, bars, and areas, to represent tabular information.
- Grammar of Graphics Renderers - Renders tabular data using a grammar of graphics approach to build charts from layered geometric shapes.
- Data-Driven Shape Generators - Translates tabular records into geometric shapes by binding data columns to visual channels like position and color.
- SVG Data Visualization - Renders complex data visualizations and geographic maps as scalable vector graphics for web browsers.
- Data Visualization Scales - Maps abstract data domains to visual properties like position and color through a system of automated scales.
- Coordinate Projections - Provides mathematical utilities to project spherical longitude and latitude coordinates into two-dimensional pixel positions.
- Categorical Data Mappings - Assigns discrete, ordinal, or nominal data to uniform visual intervals to prevent interpolation on chart axes.
- Color Schemes - Encodes data magnitudes and categories using sequential, diverging, or categorical color palettes.
- Tabular Data Pre-processing - Derives new data values on-the-fly through operations like binning and rolling averages to prepare data for visualization.
- GeoJSON Renderers - Renders geographic features using the GeoJSON specification and projects them into visual map space.
- Composable Mark Extensions - Allows creation of specialized chart types by combining primitive geometric marks or adding custom drawing logic.
- Geographic Projections - Uses mathematical projections to transform spherical longitude and latitude into two-dimensional pixel space.
- Geographic Data Mapping - Projects geographic coordinates and GeoJSON objects into visual maps to represent spatial trends.
- Data Transformation Functions - Allows the definition of user-specified functions to derive custom data, indexes, or channels before rendering.
- Server-Side Chart Image Exports - Produces SVG or PNG representations of visualizations in server-side environments by serializing the plot output.
- Display Data Transformations - Groups, bins, or stacks data points to resolve overplotting and highlight patterns before rendering visual marks.
- Small Multiples Generation - Partitions data by categorical or ordinal values to repeat a plot across multiple sub-grids for direct comparison.
- Small Multiples Visualizations - Creates grids of similar charts to enable direct comparison between different categories or subsets of data.
- Standalone Scale Generation - Generates independent scale objects that can be reused across different visualizations to maintain consistent mapping.
