Customizable questionsets and web app

[suung]

Pull Request: Integrate Question Set Functionality and Enhance Analysis Workflow

Summary

This pull request introduces significant enhancements to the Climate+Tech Sustainability Report Analytics App. The key changes include the integration of question set functionality, improved directory management, and enhanced analysis workflow. These updates aim to provide a more flexible and robust tool for sustainability report analysis.

Changes

1. Question Set Integration:

   • Added support for multiple question sets, allowing users to select and customize question sets for analysis.
   • Introduced a new question_sets directory to store YAML files defining different question sets.
   • Updated code/app.py to load and use the selected question set during analysis.

2. Directory Management:

   • Defined and created necessary directories (input, output, data, question_sets) if they do not exist.
   • Updated paths in code/app.py to dynamically use directories based on the selected question set.

3. Enhanced Analysis Workflow:

   • Modified code/app.py to include the relative path to app.py when running the analysis script.
   • Improved the handling of uploaded reports and question sets in code/web.py.
   • Added functionality to generate and download CSV files containing analysis results.
   • Implemented radar chart visualization for comparing scores across reports.

4. Environment Variable Management:

   • Added .env file support to manage environment variables securely.
   • Updated .gitignore to exclude sensitive directories and files (input/, output/, data/, .env).

5. Code Refactoring and Cleanup:

   • Refactored code/document.py, code/reader.py, and code/user_qa.py to remove hardcoded values and use environment variables.
   • Introduced utility functions in code/utils.py for loading configurations, structuring data, and generating CSV files.
   • Added a new code/config.py to manage configuration loading.

6. Documentation and Comments:

   • Added comments and documentation to improve code readability and maintainability.
   • Updated the Streamlit app to include links to relevant research papers and the Climate+Tech website.

Files Added/Modified

• .env.sample
• .gitignore
• code/app.py
• code/cfg.py
• code/config.py
• code/document.py
• code/reader.py
• code/user_qa.py
• code/utils.py
• code/web.py
• question_sets/default.yaml

How to Test

1. Clone the repository and switch to the branch for this pull request.
2. Create a .env file based on the .env.sample and add the necessary environment variables.
3. Run the Streamlit app using streamlit run code/web.py.
4. Upload a sustainability report and select a question set to run the analysis.
5. Verify that the analysis completes successfully and the results are available for download as a CSV file.
6. Check the radar chart visualization for comparing scores across reports.

Additional Notes

• Ensure that the OpenAI API key is set in the .env file for the analysis to work correctly.
• Review the changes to the directory structure and paths to avoid any potential issues with file management.

---

@pr (Diff with Main Branch)