2D data — rows and columns — as file. the native format for datasets, time series, grids, rankings, and measurements in the cybergraph
source format: CSV, TSV — any rectangular data with a header row and consistent columns
rendering
csv source → parse → column type inference → virtualized grid → GPU text cells
only the visible slice renders regardless of row count. a table file with 10M rows renders as fast as one with 10. column types (number, text, date, address) infer automatically and render with appropriate alignment and formatting. sortable. filterable. resizable columns
in the cybergraph
table is the native format of the knowledge economy. quantitative knowledge lives as table files — and quantitative knowledge is what earns
types of table files: karma ledgers, cyberank scores, experimental data, genomic expression matrices, financial time series, sensor streams, trial results, population statistics, astronomical catalogs, election results, market orders, protein interaction networks, benchmark scores
a table file is often the most linked type in scientific domains: the data that other papers cite is a table. the dataset that underlies a finding is a table. the cybergraph makes data citation first-class — link to the table, not to the paper that contains it
properties
- losslessly portable — CSV is the most universal data format. any tool can read it
- column semantics emerge from topology — the cybergraph discovers what a column means by seeing how files with similar columns cluster
- streaming-capable — large table files load progressively. the robot renders partial data while the rest transfers
- composable as state — a component file binds a table file as its data source, rendering interactive charts, forms, and dashboards from live table data
relation to other languages
table carries the numbers. formula carries the equations that model those numbers. text carries the interpretation of those numbers. vector visualizes them. together they are the complete scientific record
see csv for the source format. see datalog for querying table files. see component for binding table data to interactive UI
discover all concepts