Machine learning utilities.
Project description
MLnext Framework
MLnext Framework is an open source framework for hardware independent execution of machine learning using Python and Docker. It provides machine learning utilities. The corresponding Python package is called mlnext-framework. MLnext Framework belongs to a solution portfolio for the Digital Factory now to realize Data collection, storage, and evaluation.
Digitalization is posing numerous challenges for production — but above all, it provides countless opportunities for increasing productivity and system availability. To ensure that you benefit from the advantages of digitalization as quickly as possible, we will provide you with needs-based support — from installing simple stand-alone solutions to comprehensive digitalization concepts.
The Digital Factory now is based on the following four fields of activity:
Data collection, storage, and evaluation
Data transportation
Data security
Data usage
The four fields of action provide you with various solutions, from data acquisition to data utilization. Each individual solution will not only be tailored to your particular requirements; the fields of action can also be combined in any way or considered individually. Regardless of which path you are taking toward the Digital Factory, we will be happy to support you during the next steps.
To help you to meet today’s digitalization challenges and implement opportunities profitably, our solutions provide the following added values:
Scalability — tailored to your requirements
Tested and validated — in our own in-house production facilities
Ready-to-use — benefit from the digital transformation today
With target-oriented consultation, we will find the right solution for your Digital Factory together. Let us take on the challenges of digitalization and leverage its opportunities together.
Installation
Install this package using pip:
pip install mlnext-framework
Modules
The MLnext Framework consists of 7 modules:
import mlnext.data as data # for data loading and manipulation
import mlnext.io as io # for loading and saving files
import mlnext.pipeline as pipeline # for data preprocessing
import mlnext.plot as plot # for data visualization
import mlnext.score as score # for model evaluation
import mlnext.anomaly as anomaly # for anomaly interpretation
import mlnext.utils as utils # for utility functions
# hint: all functions can also be accessed from the root module
import mlnext
Release Information
0.4.0 (2023-01-27)
Changes
Updated the signature and docstrings of mlnext.plot.plot_signals.
Refactored mlnext.score.eval_sigmoid.
Removed roc_auc from mlnext.score.eval_metrics.
Fixed inconsistent typing import.
Removed neg_label from score.apply_threshold.
Features
Added mlnext.pipeline.ClippingMinMaxScaler.
Added mlnext.recall_anomalies and integrations for mlnext.eval_metrics, mlnext.pr_curve.
Deprecations
Deprecated y_true in favor of y in mlnext.pr_curve.
score.apply_threshold superceedes the functionality of score.eval_sigmoid and is removed in 0.6.
Copyright (c) 2022 Phoenix Contact Electronics GmbH
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for mlnext_framework-0.4.0-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 49b7d1c4e56117e34abf79ac2fb73e02b2999c8c0da5712d13dad2c6f240de32 |
|
MD5 | 8c261567a953620e704d4efd318d1496 |
|
BLAKE2b-256 | 5ea977597d9fa7d81dcd247f02c72118b852749cb88f6a768fd9424ff3a3b7ce |