AutoStatLib

Create a fully-functional mini-app called 'DataInsightPro' using the Python package 'AutoStatLib'. This app will serve as a user-friendly interface for conducting basic statistical analyses on datasets. Users should be able to upload their dataset (CSV format), select columns of interest, and then perform various statistical operations such as mean, median, mode, standard deviation, variance, correlation coefficients, and histograms. Additionally, the app should provide visualizations for the data through bar charts, pie charts, and scatter plots to help users understand the relationships between variables. 

The app should have a clean, intuitive UI built with Streamlit or Dash, allowing users to interactively explore their data. Upon uploading a dataset, users should see summary statistics for all numerical columns, and they should be able to choose specific columns to analyze further. The results of each analysis should be displayed in a structured format, including both numerical summaries and visual representations.

Utilize 'AutoStatLib' for its core statistical functions. For instance, use it to calculate descriptive statistics like mean, median, etc., and to generate correlation matrices. Integrate any visualization libraries like Matplotlib or Seaborn to display the data visually. Ensure that the app handles errors gracefully, such as incorrect file formats or missing data, providing clear error messages to the user.

As an additional feature, implement a simple machine learning model (e.g., linear regression) using scikit-learn, which predicts one variable based on another, using the data provided by the user. Use 'AutoStatLib' to evaluate the performance of the model by calculating metrics like R-squared and Mean Squared Error. Finally, ensure the app is well-documented, with clear instructions on how to install dependencies and run the application.