Splitting Datasets

---
name: splitting-datasets
description: 'Process split datasets into training, validation, and testing sets for
  ML model development. Use when requesting "split dataset", "train-test split", or
  "data partitioning". Trigger with relevant phrases based on skill purpose.

  '
allowed-tools: Read, Write, Edit, Grep, Glob, Bash(cmd:*)
version: 1.0.0
license: MIT
tags:
- ai
- testing
- ml
compatibility: Designed for Claude Code, also compatible with Codex and OpenClaw
---
# Dataset Splitter

Split datasets into training, validation, and testing sets with configurable ratios and stratification options.

## Overview

This skill automates the process of dividing a dataset into subsets for training, validating, and testing machine learning models. It ensures proper data preparation and facilitates robust model evaluation.

## How It Works

1. **Analyze Request**: The skill analyzes the user's request to determine the dataset to be split and the desired proportions for each subset.
2. **Generate Code**: Based on the request, the skill generates Python code utilizing standard ML libraries to perform the data splitting.
3. **Execute Splitting**: The code is executed to split the dataset into training, validation, and testing sets according to the specified ratios.

## When to Use This Skill

This skill activates when you need to:
- Prepare a dataset for machine learning model training.
- Create training, validation, and testing sets.
- Partition data to evaluate model performance.

## Examples

### Example 1: Splitting a CSV file

User request: "Split the data in 'my_data.csv' into 70% training, 15% validation, and 15% testing sets."

The skill will:
1. Generate Python code to read the 'my_data.csv' file.
2. Execute the code to split the data according to the specified proportions, creating 'train.csv', 'validation.csv', and 'test.csv' files.

### Example 2: Creating a Train-Test Split

User request: "Create a train-test split of 'large_dataset.csv' with an 80/20 ratio."

The skill will:
1. Generate Python code to load 'large_dataset.csv'.
2. Execute the code to split the dataset into 80% training and 20% testing sets, saving them as 'train.csv' and 'test.csv'.

## Best Practices

- **Data Integrity**: Verify that the splitting process maintains the integrity of the data, ensuring no data loss or corruption.
- **Stratification**: Consider stratification when splitting imbalanced datasets to maintain class distributions in each subset.
- **Randomization**: Ensure the splitting process is randomized to avoid bias in the resulting datasets.

## Integration

This skill can be integrated with other data processing and model training tools within the Claude Code ecosystem to create a complete machine learning workflow.

## Prerequisites

- Appropriate file access permissions
- Required dependencies installed

## Instructions

1. Invoke this skill when the trigger conditions are met
2. Provide necessary context and parameters
3. Review the generated output
4. Apply modifications as needed

## Output

The skill produces structured output relevant to the task.

## Error Handling

- Invalid input: Prompts for correction
- Missing dependencies: Lists required components
- Permission errors: Suggests remediation steps

## Resources

- Project documentation
- Related skills and commands