Metadata-Version: 2.1
Name: wtpcc
Version: 0.1.0
Summary: A lightweight function for normal behavior modeling for wind turbine SCADA data.
Author-Email: fromZeroToInf <syh_82@outlook.de>
License: MIT
Requires-Python: >=3.12
Requires-Dist: pandas>=3.0.2
Requires-Dist: numpy>=2.4.4
Requires-Dist: psutil>=7.2.2
Requires-Dist: joblib>=1.5.3
Requires-Dist: tqdm>=4.67.3
Requires-Dist: pyarrow>=23.0.1
Description-Content-Type: text/markdown

# Wind-Turbine-SCADA-Data-Power-Curve-Cleaning (wtpcc)
`wtpcc` is a lightweight Python package for power-curve-based cleaning of wind turbine SCADA data.

The package filters SCADA observations by comparing measured power values against an interpolated reference power curve. In addition to point-wise filtering, it supports simple window-based removal of neighboring samples around out-of-band observations. It also provides file-based batch processing for CSV and Parquet files with optional parallel execution.

## Features

- power-curve-based filtering for SCADA data
- linear interpolation of a reference power curve via `numpy.interp`
- configurable power margin around the reference power curve
- minimum wind-speed threshold
- window-based filtering using convolution.
- batch processing of CSV and Parquet files
- optional parallel processing with automatic job estimation based on available RAM

## Installation
```bash
pip install wtpcc
```

# From Source
```bash
pip install .
or
pdm install
or 
pip install -e .
```

