vaex-dataframes

Out-of-core DataFrame for billion-row data via lazy evaluation and memory-mapped files. Use when data exceeds RAM (10 GB–TB) for fast aggregation, filtering, virtual columns, and visualization without loading. Supports HDF5, Arrow, Parquet, CSV with cloud (S3, GCS, Azure). Built-in ML transformers (scaling, PCA, K-means). In-memory: polars; distributed: das…

Independent GitHub snapshot · Automated classification · Original documentation governs