coding-data-science

Purpose

This skill allows OpenClaw to handle data science tasks using Python, R, and Julia, focusing on machine learning, data wrangling, visualization, statistical modeling, and notebook execution. It integrates these languages to enable seamless code execution and analysis.

When to Use

Use this skill when users need to process datasets, build ML models, or perform statistical analysis. Apply it for tasks like exploratory data analysis in Jupyter notebooks, training models with scikit-learn in Python, or visualizing data with ggplot2 in R. Avoid it for non-data tasks like web development.

Key Capabilities

• Execute Python code for data manipulation (e.g., via pandas) and ML (e.g., scikit-learn).
• Run R scripts for statistical modeling (e.g., using glm) and visualization (e.g., ggplot2).
• Support Julia for high-performance computing in ML (e.g., Flux.jl).
• Handle notebook formats like Jupyter for interactive sessions.
• Integrate with libraries: Python's numpy for arrays, R's dplyr for data wrangling, Julia's DataFrames.jl.
• Process large datasets with memory-efficient methods, such as Python's dask for parallel computing.

Usage Patterns

To use this skill, invoke OpenClaw via CLI or API, specifying the language and code. Always set the environment variable $OPENCLAW_API_KEY for authentication. For example, prefix commands with the skill ID: openclaw execute --skill coding-data-science. Use JSON config files for multi-step workflows, e.g., {"lang": "python", "code": "import pandas as pd"}. Chain tasks by piping outputs, like running a Python script that generates data for an R visualization.