idsw 1.2.6

Full workflow for ETL, statistics, and Machine learning modelling of (usually) time-stamped industrial facilities data.

Industrial Data Science Workflow

Industrial Data Science Workflow: full workflow for ETL, statistics, and Machine learning modelling of (usually) time-stamped industrial facilities data.

Not only applicable to monitoring quality and industrial facilities systems, the package can be applied to data manipulation, characterization and modelling of different numeric and categorical datasets to boost your work and replace tradicional tools like SAS, Minitab and Statistica software.

• Check the project Github: https://github.com/marcosoares-92/IndustrialDataScienceWorkflow

• Check our Steel Industry Simulator on: https://github.com/marcosoares-92/steelindustrysimulator

  • The Ideal Tool for Process Improvement, and Data Collection, Analyzing and Modelling Training.
  • User interface available in:

  https://colab.research.google.com/github/marcosoares-92/steelindustrysimulator/blob/main/steelindustry_digitaltwin.ipynb

Authors:

• Marco Cesar Prado Soares, Data Scientist Specialist at Bayer (Crop Science)

  • [e-mail]

• Gabriel Fernandes Luz, Senior Data Scientist

  • [e-mail]

• If you cannot install the last version from idsw package directly from PyPI using pip install idsw:

1. Open the terminal and:

Run:

git clone "https://github.com/marcosoares-92/IndustrialDataScienceWorkflow" 

to clone all the files (you could also fork them).

2. Go to the directory called idsw.
3. Now, open the Python terminal and:

Navigate to the idsw folder to run:

pip install .

• You can use command cd "...\idsw", providing the full idsw path to navigate to it. Alternatively, run pip install ".\*.tar.gz" in the folder terminal.

After cloning the directory, you can also run the package without installing it:

1. Copy the whole idsw folder to the working directory where your python or jupyter notebook file is saved.

• There must be an idsw folder on the python file directory.

2. In your Python file:

Run the command or run a cell (Jupyter notebook) with:

from idsw import *

for importing all idsw functions without the alias idsw; or:

import idsw

to import the package with the alias idsw.

History

1.2.0

Fixed

• Deprecated structures

Added

• New functionalities added.

Reshape of project design.

• New division into modules and new names for functions and classes.

Removed

• Removed support for Python < 3.7

1.2.1

Fixed

• Setup issues.

1.2.2

Fixed

• Setup issues: need for rigid and specific versions of the libraries.

1.2.3

Fixed

• Setup issues.

1.2.4

Fixed

• Import bugs.
• Introduced function for Excel writing.

1.2.5

Fixed

• Matplotlib export figures bugs.
• 'quality' argument is no longer supported by plt.savefig function (Matplotlib), so it was removed.
• This modification was needed for allowing the correct functioning of the steelindustrysimulator, which is based on idsw.
• Check simulator project on: https://github.com/marcosoares-92/steelindustrysimulator
  • The Ideal Tool for Process Improvement, and Data Collection, Analyzing and Modelling Training.

1.2.6

Fixed

• Export of figures generated a message like with '{new_file_path}.png.png'. Fixed to '{new_file_path}.png'.