Skyworker Processor: {processor_name}

Template repository for the Skyworker AI Processor

Requirements

• Python >= 3.6
• pip
• venv (recommended)

Usage

• Clone this repository
• Update todos in setup.py file
• Update todos in skyworker_processor/__init__.py file
• Define your logic in skyworker_processor/core/processor.py module's process function

The return value of skyworker_processor.process() function should be the pandas.DataFrame object with following columns:

• measurement - Measurement Name
• value - Calculated Value

and the index of each row should be the unique timestamp.

Example

import pandas as pd
from datetime import datetime

time_series = [
    {
        "timestamp": 1628150400,
        "measurement": "Electricity Price",
        "value": 0.23,
    },
    {
        "timestamp": 1628236800,
        "measurement": "Electricity Price",
        "value": 0.25,
    }
]

data_frame = pd.DataFrame(time_series)
# Set "timestamp" column as index.
data_frame.index = data_frame["timestamp"]
# Compose "time" column from the "timestamp"
data_frame["time"] = data_frame["timestamp"].map(datetime.fromtimestamp)
del data_frame["timestamp"]

Build & Publish

Build and Publish is a part of the CI/CD pipeline, but you can publish your package manually:

# Check the package.
$ python setup.py check

# Build the package.
$ python setup.py sdist

# Publish the package.
$ twine upload dist/*

Changelog

Please see CHANGELOG for details.

Authors

• Temuri Takalandze - Author and Maintainer

License

Copyright © 2021 Omedia.
Released under the Other/Proprietary License.