Datasets:

THULab
/

coupled_msd

pretty_name: Coupled Mass-Spring-Damper (TsFile)
modality: timeseries
authors: dany-l-23
task_categories:
  - time-series-forecasting
size_categories:
  - 1M<n<10M
tags:
  - tsfile
  - timeseries
  - modality:timeseries
  - format:tsfile
configs:
  - config_name: default
    data_files:
      - split: train
        path: coupled_msd_*.tsfile

Coupled Mass-Spring-Damper (TsFile)

This dataset is an Apache TsFile conversion of dany-l-23/coupled-msd.

Modalities: Time-series.

Overview

Input–output data of a coupled mass-spring-damper system with a nonlinear force profile (generated with statesim).
200 simulation trajectories of length 7500, split 120 train / 20 validation / 60 test.
u_1 is the input, y_1 the output, and x_1..x_8 the internal states.
The train/validation/test split is preserved as three tsfiles; each trajectory is a device via the traj_id TAG.
Converted observations: 1,500,000 rows across 3 TsFile file(s)
Source format: csv

TsFile schema

Time — source t (seconds), converted to INT64 milliseconds.

Column	Role	Type	Meaning
`Time`	TIME	INT64 (ms)	sample timestamp
`traj_id`	TAG	STRING	trajectory id
`u_1`	FIELD	FLOAT	input
`y_1`	FIELD	FLOAT	output
`x_1`	FIELD	FLOAT	state 1
`x_2`	FIELD	FLOAT	—
`x_3`	FIELD	FLOAT	—
`x_4`	FIELD	FLOAT	—
`x_5`	FIELD	FLOAT	—
`x_6`	FIELD	FLOAT	—
`x_7`	FIELD	FLOAT	—
`x_8`	FIELD	FLOAT	—

Conversion notes

Split preserved: coupled_msd_train.tsfile, coupled_msd_validation.tsfile, coupled_msd_test.tsfile.
t consumed into Time; all 10 measurement/state columns kept as FLOAT FIELDs.

Source & license

Original dataset: https://huggingface.co/datasets/dany-l-23/coupled-msd
Author / publisher: dany-l-23
License: mit

Usage

Install the Apache TsFile Python SDK (pip install tsfile) and read a converted file:

from pathlib import Path
from tsfile import TsFileReader

path = Path("coupled_msd_test.tsfile")
with TsFileReader(str(path)) as reader:
    schemas = reader.get_all_table_schemas()
    print("tables:", list(schemas))
    table_name = next(iter(schemas))
    table = schemas[table_name]
    columns = [column.get_column_name() for column in table.get_columns()]
    print("columns:", columns)
    field_names = [
        column.get_column_name()
        for column in table.get_columns()
        if column.get_column_name() not in {"Time", "time"}
    ]
    if field_names:
        with reader.query_table(table_name, field_names[:3], batch_size=1024) as result:
            batch = result.read_arrow_batch()
            if batch is not None:
                print(batch.to_pandas().head())