This is an experimental software providing a point and click user interface for generating python code for machine learning (ML) pipelines. At this early stage, the features are very limited. We are actively working on building new features to get the first proof of concept version working. The goal of this software is to allow non-programmers "write" ML pipeline code.
The instructions below assume that python3 and pip3 refer to Python 3.x and Pip 3.x. Depending on how things are configured, you may need to use python and pip commands instead of python3 and pip3. Additionally, Windows users would benefit from installing gitbash for windows, which will allow them to run bash commands (as used in these instructions).
- Create a virtual environment: python3 -m venv venv
- Activate virtual environment:
- Mac or Linux: source venv/bin/activate
- Windows: source venv/Scripts/activate
- Install dependencies: pip3 install -r requirements.txt
- Activate virtual environment each time you use this code:
- Mac or Linux: source venv/bin/activate
- Windows: source venv/Scripts/activate
export PYTHONPATH=$(pwd)
- python3 driver.py <INPUT_CSV_FILE>
This code will process the input csv file, remove rows with missing values, and will output data summary. Generated python code will be printed to the terminal.
- python3 flask_app/flask_main.py
If all goes well, you should see the following in your terminal:
* Serving Flask app 'flask_main'
* Debug mode: off
WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
* Running on http://127.0.0.1:5000
Press CTRL+C to quit
Point your web browser to http://127.0.0.1:5000 to access the user interface of this software.
- pytest
- deactivate