Ray Datasets API

• Input/Output
  • Synthetic Data
  • Python Objects
  • Parquet
  • CSV
  • JSON
  • Text
  • Images
  • Binary
  • TFRecords
  • Pandas
  • NumPy
  • Arrow
  • MongoDB
  • SQL Databases
  • Dask
  • Spark
  • Modin
  • Mars
  • Torch
  • Hugging Face
  • TensorFlow
  • WebDataset
  • Datasource API
  • Partitioning API
  • MetadataProvider API
• Dataset API
  • Constructor
  • Basic Transformations
  • Sorting, Shuffling, Repartitioning
  • Splitting and Merging Datasets
  • Grouped and Global Aggregations
  • Converting to Pipeline
  • Consuming Datasets
  • I/O and Conversion
  • Inspecting Metadata
  • Execution
  • Serialization
  • Internals
• DatasetIterator API
  • ray.data.DatasetIterator.iter_batches
  • ray.data.DatasetIterator.iter_torch_batches
  • ray.data.DatasetIterator.to_tf
  • ray.data.DatasetIterator.stats
• DatasetPipeline API
  • Constructor
  • Basic Transformations
  • Sorting, Shuffling, Repartitioning
  • Splitting DatasetPipelines
  • Creating DatasetPipelines
  • Consuming DatasetPipelines
  • I/O and Conversion
  • Inspecting Metadata
• ExecutionOptions API
  • Constructor
  • Resource Options
• GroupedDataset API
  • Constructor
  • Computations / Descriptive Stats
  • Function Application
  • Aggregate Function
• DatasetContext API
  • Constructor
  • Get DatasetContext
• Data Representations
  • Block API
  • Batch API
  • Row API
  • Tensor Column Extension API
• (Experimental) RandomAccessDataset API
  • Constructor
  • Functions
• Utility
  • ray.data.set_progress_bars
• API Guide for Users from Other Data Libraries
  • For Pandas Users
  • For PyArrow Users