Contents Menu Expand Light mode Dark mode Auto light/dark, in light mode Auto light/dark, in dark mode Skip to content
StandardE2E 0.0.5
Logo
StandardE2E 0.0.5

Getting Started

  • Quickstart
  • Overview
  • Supported Datasets

API Reference

  • API Reference
    • Unified Dataset
    • Data Structures
      • StandardFrameData
      • FrameIndexData
      • TransformedFrameData
      • TransformedFrameDataBatch
      • Trajectory
      • BatchedTrajectory
      • CameraData
      • LidarData
      • LidarPointCloud
      • BatchedLidarPointCloud
      • Detection3D
      • FrameDetections3D
      • BatchedFrameDetections3D
      • HDMap
      • MapElement
    • Enums
    • Caching and Processing
      • Caching Core
        • SourceDatasetProcessor
        • SourceDatasetConverter
      • Caching Adapters
        • AbstractAdapter
        • PanoImageAdapter
        • CamerasIdentityAdapter
        • IntentIdentityAdapter
        • FutureStatesIdentityAdapter
        • PastStatesIdentityAdapter
        • PreferenceTrajectoryAdapter
        • Detections3DIdentityAdapter
        • LidarAdapter
        • LidarBEVAdapter
        • HDMapBEVAdapter
      • Caching Segment Context
        • SegmentContextAggregator
        • FuturePastStatesFromMatricesAggregator
        • FutureDetectionsAggregator
      • Caching Source Datasets
        • WaymoE2EDatasetProcessor
        • WaymoE2EDatasetConverter
        • WaymoPerceptionDatasetProcessor
        • WaymoPerceptionDatasetConverter
        • Av2SensorDatasetProcessor
        • Av2SensorDatasetConverter
        • Av2LidarDatasetProcessor
        • Av2LidarDatasetConverter
        • NavsimDatasetProcessor
        • NavsimDatasetConverter
    • Dataset Utilities
      • FrameLoader
      • standard_e2e.dataset_utils.frame_loader.create_frame_loaders_from_config
      • FrameSelector
      • standard_e2e.dataset_utils.selector.create_frame_selector_from_config
    • Modality Defaults
      • ModalityDefaults
      • IntentDefaults
      • PreferredTrajectoryDefaults
      • LidarPointCloudDefaults
      • LidarBEVDefaults
      • HDMapBEVDefaults
    • Augmentation
      • FrameAugmentation
      • IdentityFrameAugmentation
      • MultipleFramesImageAugmentation
      • TrajectoryResampling
    • Indexing
      • IndexDataGenerator
      • IndexFilter
      • FrameFilterByBooleanColumn
      • FrameFilterByTimeDelta

Tutorials

  • đź““ Introduction Tutorial
  • đź““ Data Containers
  • đź““ Multi-Dataset Training
  • đź““ Custom Adapters

Project Links

  • GitHub
  • Discord
  • PyPI
Back to top
View this page

NavsimDatasetProcessor¶

class standard_e2e.caching.src_datasets.navsim.NavsimDatasetProcessor(common_output_path, split, index_data_generator=None, adapters=None, context_aggregators=None, maps_root_path=None)[source]¶

Bases: SourceDatasetProcessor

Processor for NAVSIM (OpenScene-v1.1 trainval).

A “frame” is one ego timestamp inside a scene pickle. Per-log state (the entire pickle) is loaded once and reused for every frame in that log; the cache key is the pickle path, so workers that hop between logs only reload at log boundaries.

HD-map handling: nuPlan map.gpkg files for the 4 OpenScene cities are loaded lazily via get_maps_api and cached per map_location string (so each worker holds at most 4 NuPlanMap instances). The maps root is resolved with the same precedence NAVSIM users expect:

  1. explicit maps_root_path constructor arg;

  2. NUPLAN_MAPS_ROOT environment variable (NAVSIM’s own convention, documented at navsim/docs/install.md);

  3. <input_path>/maps derived from the converter’s input_path at frame-prep time (matches OpenScene-v1.1’s on-disk layout).

If none of the three resolves to an existing directory, HD-map extraction is skipped and a warning is logged once.

Parameters:
  • common_output_path (str)

  • split (str)

  • index_data_generator (IndexDataGenerator | None)

  • adapters (list[AbstractAdapter] | None)

  • context_aggregators (list[SegmentContextAggregator] | None)

  • maps_root_path (Optional[str])

DATASET_NAME = 'navsim'¶
property allowed_splits: list[str]¶

Return the list of allowed splits for the dataset.

property context_aggregators¶
property dataset_name: str¶

Return the name of the dataset.

property inner_path: str¶

Return the inner path relative to the common output path.

needs_attr(attr)¶

Whether at least one registered adapter reads this StandardFrameData field. Used by per-dataset processors to skip expensive modality builds (cameras, lidar, hd_map, detections, …) when no adapter would consume them. True when attr is in the consumed-attrs union, plus a hard-coded special case: the identifier / index fields are always treated as needed since they are required for the cache + index regardless of adapter chain.

Return type:

bool

Parameters:

attr (StandardFrameDataField)

property output_path: str¶

Return the output path for the processed dataset.

process_frame(raw_frame_data)¶
Return type:

tuple[TransformedFrameData, FrameIndexData]

Parameters:

raw_frame_data (Any)

process_frame_and_save_data(raw_frame_data)¶

Process a single frame of raw data, save the processed frame data to disk, and return the corresponding FrameIndexData.

Return type:

FrameIndexData

Parameters:

raw_frame_data (Any)

property specific_output_path: str¶

Return the specific output path for the dataset.

property split: str¶

Return the dataset split.

Next
NavsimDatasetConverter
Previous
Av2LidarDatasetConverter
Copyright © 2026, Stepan Konev and contributors
Made with Sphinx and @pradyunsg's Furo
On this page
  • NavsimDatasetProcessor
    • NavsimDatasetProcessor
      • NavsimDatasetProcessor.DATASET_NAME
      • NavsimDatasetProcessor.allowed_splits
      • NavsimDatasetProcessor.context_aggregators
      • NavsimDatasetProcessor.dataset_name
      • NavsimDatasetProcessor.inner_path
      • NavsimDatasetProcessor.needs_attr()
      • NavsimDatasetProcessor.output_path
      • NavsimDatasetProcessor.process_frame()
      • NavsimDatasetProcessor.process_frame_and_save_data()
      • NavsimDatasetProcessor.specific_output_path
      • NavsimDatasetProcessor.split