---
license: apache-2.0
task_categories:
- robotics
tags:
- LeRobot
- tsfile
- time-series
- robotics
- timeseries
modality:
- timeseries
configs:
- config_name: default
  data_files:
  - split: train
    path: data/roarm_m3_test.tsfile
---

# roarm_m3_test

This dataset is converted from
[`dudulrx0601/roarm_m3_test`](https://huggingface.co/datasets/dudulrx0601/roarm_m3_test)
to Apache TsFile format.

Modalities: Time-series. The original dataset also contains laptop and phone
camera video streams, which remain in the source Hugging Face dataset and are
not included in this converted repository.

## Source Dataset

- Original dataset: [`dudulrx0601/roarm_m3_test`](https://huggingface.co/datasets/dudulrx0601/roarm_m3_test)
- Format: LeRobot v2.0
- Robot type: `unknown`
- License: `apache-2.0`
- Scale: 50 episodes, 4,739 frames, 1 task (`pick`), 30 fps
- Original data path: `data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet`
- Original video path: `videos/chunk-{episode_chunk:03d}/{video_key}/episode_{episode_index:06d}.mp4`

## Converted Files

- TsFile data: `data/roarm_m3_test.tsfile`
- Mirrored metadata: `meta/`
- Rows in TsFile: 4,739
- Time precision: milliseconds

## Schema

`Time` is synthesized as `round(timestamp * 1000)` and restarts per episode.

TAG columns:

- `episode_index`
- `task_index`

FIELD columns:

- `frame_index`
- `sample_index` renamed from source `index`
- `next_done` renamed from source `next.done`
- `observation_state_0` through `observation_state_5`, flattened from `observation.state[6]`
- `action_0` through `action_5`, flattened from `action[6]`

Flattened vector values are stored as single-precision FLOAT fields.

## Conversion Notes

- The source `timestamp` column is not retained as a separate field because it
  is redundant with `Time / 1000` seconds.
- Source video streams `observation.images.laptop` and
  `observation.images.phone` are not uploaded here. Use the original dataset
  videos:
  https://huggingface.co/datasets/dudulrx0601/roarm_m3_test/tree/main/videos
- `meta/info.json` is mirrored from the source and updated to describe the
  converted TsFile schema, original video path, row count, TAG columns,
  flattened features, renamed fields, and video policy.
- Aside from the redundant `timestamp` field and omitted video streams, no rows
  or numeric time-series fields are intentionally dropped.

## Read Example

```python
# Read data/roarm_m3_test.tsfile with the Apache TsFile Java/Python SDK.
# Query one episode with a predicate such as:
# WHERE episode_index = 0
```