Climate Question Processing Tools

This directory contains tools for batch processing of climate-related questions, including generation, visualization, and execution. These tools are designed to work together in a pipeline to generate climate questions, validate them, and process them using the Climsight system.

Tools Overview

1. Question Generator (question_generator.py)

Generates climate-related questions with geographic locations and validates that the locations are on land.

Key Features:

• Generates questions by theme and type (general/specific)
• Validates question locations against land/water boundaries
• Can validate existing question files
• Saves questions to JSON files

Usage:

# Generate new questions
python question_generator.py

# Validate existing questions file
python question_generator.py questions_01.json

2. Question Map (question_map.py)

Interactive visualization tool for exploring climate questions on a map.

Key Features:

• Displays questions on an interactive map
• Color-codes questions by theme and type
• Supports filtering by theme and source file
• Shows summary statistics

Usage:

# Run the Streamlit app
streamlit run question_map.py

3. Question Runner (question_runner.py)

Processes batches of questions through the Climsight system to generate answers.

Key Features:

• Processes questions in sequence
• Saves progress to allow resuming interrupted runs
• Supports different LLM models
• Generates detailed output files

Usage:

# Process questions from a file
python question_runner.py --questions_file Q_1.json --llm_model gpt-4.1-nano

Batch Processing Workflow

1. Generate Questions (optional):

   cd /path/to/climsight
   cd sequential
   python question_generator.py
   

2. Visualize Questions (optional):

   cd /path/to/climsight
   cd sequential
   streamlit run question_map.py
   

3. Process Questions:

   cd /path/to/climsight
   cd sequential
   python question_runner.py --questions_file Q_1.json --llm_model gpt-4.1-nano
   

File Structure

• Generated question files: questions_XX.json
• Output answers: answers_output.json
• Temporary progress: temporal_output.json
• Configuration: config.yml (in project root)

Notes

• All scripts should be run from the project sequential directory
• Ensure you have the required environment variables set (e.g., OPENAI_API_KEY)