uwdata/arquero v2.2.0

on GitHub

Changes from v2.1.3:

• Update table() constructor to accept Map objects and [[key, value], ...] entry arrays as input.
• Update fromCSV() parsing to ensure consistent types across column values.
• Add autoMax option to fromCSV() to control how many initial rows are used for type inference.
• Add assign() method to bind columns from multiple tables into a new table.
• Add slice() verb to extract contiguous rows from table groups.
• Add frac() helper to generate expressions for per-group row counts. Useful as input to the sample() verb.
• Add column iterator support. (thanks @mbostock!)
• Add column factory mechanism, export as internal.columnFactory() getter/setter.
• Add exports for internal.Transformable and internal.Verb classes.
• Add internal performance benchmark scripts.
• Refactor: Consolidate default null / missing value.
• Refactor: Consolidate internal loop unrolling utilities.
• Refactor: Reorganize directory structure.