/autolit

Streamlining explanatory data analysis and modeling of tabular information. Wrapped in a streamlit app.

Primary LanguagePythonMIT LicenseMIT

autolit v0.01 Streamlit App

Streamlining explanatory data analysis and machine-learning of tabular information, and wrapping it in a streamlit app.

Click streamlit badge above to use app. Streamlit App

Work flow of app

  • Upload Data
    • Choose where to get your data.
      • Toy data sets available:
        • Iris dataset for classification
        • Boston housing dataset for regression
      • Upload your own from local machine
      • Insert a link to data set on the internet
    • Confirm file type and import
      • Supports both xls and csv file types
  • Explore Data
    • Observe interesting plots
      • Ranked by skew for distribution plots
      • Randked by correlation for scatterplots
    • Count data entries and missing values
    • Correlation matrix
    • Optional boxplots and countplots for further examination
  • Modeling
    • Construct pipeline to predict on data
    • Plot feature importance
    • Plot learning curve
│   .gitignore
│   Dockerfile
│   LICENSE
│   main.py
│   README.md
│   requirements.txt
│   
│       
├───autolit
│          alt_plotter.py
│          autopipe.py
│          data_reader.py
│          file_importer.py
│          lr_plot.py
│          slide.py
│          sns_plotter.py
│      
│ 
│
└───src
        script.js
        slide.html
        style.css