Datasets:

THULab
/

azure

pretty_name: Azure Predictive Maintenance (TsFile)
modality: timeseries
authors: odysseywt
task_categories:
  - time-series-forecasting
size_categories:
  - 100K<n<1M
tags:
  - tsfile
  - timeseries
  - modality:timeseries
  - format:tsfile
configs:
  - config_name: default
    data_files:
      - split: train
        path: azure.tsfile

Azure Predictive Maintenance (TsFile)

This dataset is an Apache TsFile conversion of odysseywt/Azure.

Modalities: Time-series.

Overview

Industrial machine telemetry for predictive maintenance (Azure demo dataset).
Voltage, rotation, pressure, and vibration readings with machine age and a failure label.
Each machine is a device identified by the machineID TAG.
Converted observations: 876,100 rows across 1 TsFile file(s)
Source format: csv

TsFile schema

Time — source datetime (datetime), converted to INT64 milliseconds.

Column	Role	Type	Meaning
`Time`	TIME	INT64 (ms)	sample timestamp
`machineID`	TAG	STRING	machine id
`volt`	FIELD	FLOAT	voltage
`rotate`	FIELD	FLOAT	rotation
`pressure`	FIELD	FLOAT	pressure
`vibration`	FIELD	FLOAT	vibration
`age`	FIELD	FLOAT	machine age
`label`	FIELD	FLOAT	failure label

Conversion notes

machineID kept as TAG; telemetry + age + label as FIELDs.

Source & license

Original dataset: https://huggingface.co/datasets/odysseywt/Azure
Author / publisher: odysseywt
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("azure.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())