README.md

---
pretty_name: "MFPT Bearing Fault (TsFile)"
modality: timeseries
authors: "odysseywt"
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: "mfpt.tsfile"
---

# MFPT Bearing Fault (TsFile)

This dataset is an Apache TsFile conversion of
[`odysseywt/MFPT`](https://huggingface.co/datasets/odysseywt/MFPT).

Modalities: Time-series.

## Overview

- Machinery Failure Prevention Technology (MFPT) bearing dataset.
- Vibration signals across conditions, bearing numbers, load conditions, and sampling rates.
- Condition/bearing/load/sampling metadata are device TAGs; `vibration_raw` is the measurement.

- Converted observations: 5,544,960 rows across 1 TsFile file(s)
- Source format: csv

## TsFile schema

- **Time** — sample index within each segment, stored as INT64 milliseconds.

| Column | Role | Type | Meaning |
|---|---|---|---|
| `Time` | TIME | INT64 (ms) | sample timestamp |
| `condition` | TAG | STRING | fault condition |
| `bearing_num` | TAG | STRING | bearing id |
| `load_condition` | TAG | STRING | load |
| `label` | TAG | STRING | label |
| `segment_idx` | TAG | STRING | — |
| `sampling_rate` | TAG | STRING | Hz |
| `segment_id` | TAG | STRING | segment id |
| `vibration_raw` | FIELD | FLOAT | amplitude |

## Conversion notes

- Metadata columns kept as TAGs; `vibration_raw` as FLOAT FIELD.
- Extraneous source columns (`features`, `bearing_type`, `filename`) dropped.

## Source & license

- Original dataset: https://huggingface.co/datasets/odysseywt/MFPT
- Author / publisher: odysseywt
- License: cc-by-4.0

## Usage

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

```python
from pathlib import Path
from tsfile import TsFileReader

path = Path("mfpt.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())
```