bring-order 1.0.1

Improved tools for data scientists

bring-order

The BringOrder tool is aimed at guiding data scientists with their analysis using custom widgets inside Jupyter Notebook.

The first step is to import and prepare data for analysis, and to identify data limitations. If the data set is loaded from a CSV file, BringOrder checks automatically if the variables are normally distributed and offers an option to check if selected variables are independent.

After the data preparation phase, there are two analysis paths that the user can choose from: hypothesis testing or explorative analysis. Testing a hypothesis includes summarizing theory for context, formulating the hypotheses, running some analysis code, and drawing a conclusion.

The explorative analysis starts with stating preconceptions and ends with evaluating the analysis with respect to those preconceptions. In between, the user should run some analysis code and write some notes about their observations. To help the user with writing code, BringOrder provides the possibility to use OpenAI's ChatGPT inside Jupyter Notebook through an API connection.

As a result, BringOrder produces a structured notebook where all the steps are documented in Markdown. The user also has the option to save a PowerPoint presentation template of the analysis and to export the notebook to a PDF file.

Documentation

• Product backlog

• Definition of Done

• Minutes of meetings

Installation

Installation from PyPi

Linux

• If you don't have Jupyter Notebook installed, install it with

    pip install notebook

• If you don't have ipywidgets installed, install it with

    pip install ipywidgets

• If you have an old version of ipywidgets installed, upgrade it with

    pip install --upgrade ipywidgets

• Install bring-order extension:

    pip install bring-order

Windows

• If you don't have PIP installed, it is recommended to use Anaconda environment

• Install Anaconda with Anaconda download

• Use Anaconda Powershell prompt instead of regular Powershell or Command prompt

• Install bring-order extension:

    pip install bring-order

MacOS

• If you don't have ipywidgets installed, install it with

    python3 -m pip install "ipywidgets"

• If you have an old version of ipywidgets installed, upgrade it with

    python3 -m pip install --upgrade ipywidgets

• Install bring-order extension:

    python3 -m pip install "bring-order"

Usage

• Open Jupyter Notebook with

    jupyter notebook

• In Jupyter Notebook execute

    from bring_order import BringOrder
    BringOrder()

A more detailed user manual can be found in the User Guide

Development and Testing

Links to development documentation

Architecture

Testing report

Suggestions for improvements

Dependencies

• Clone the project and install dependencies in the main folder

    poetry install

• Navigate to main folder and go to the virtual environment

    poetry shell

Testing

Run unit tests

    pytest tests

or use Invoke

    invoke tests

or generate Covegare report

    invoke coverage

Run style check

    pylint bringorder

or use Invoke

    invoke lint

Run robot tests

    ./run_robot_tests.sh

or use Invoke

    invoke robottests

Run pylint, unit tests, and robot tests at once

invoke alltests