Apply type casting, column ordering, and deterministic splits to finalize datasets for modeling.
npx skills add https://github.com/thbraet/claude-template --skill format-data
Please help me install this Agent Skill. Skill: format-data Source: https://github.com/thbraet/claude-template/tree/main/skills/format-data Command: npx skills add https://github.com/thbraet/claude-template --skill format-data
Format-data resolves the need to finalize a combined dataset for modeling by applying final formatting transformations such as type casting, column ordering, and deterministic splits.
Format the integrated dataset (3.4) into a modeling-ready schema and generate the accompanying 3.5-format-data notebook and summary.
High-intent search queries and answers about installing and using this skill.
Dataset formatting for modeling applies type casting, column ordering, and deterministic train/validation/test partitioning to finalize a combined dataset. It enforces a structured, version-controllable output pipeline ready for machine learning algorithms.
CRISP-DM 3.5 data preparation is the final formatting phase before modeling. You need it to apply deterministic splits, type casting, and column ordering to raw data, transforming it into a modeling-ready schema with reproducible loading instructions.
To format raw data into a modeling-ready schema, apply type casting, column ordering, renaming, and encoding transformations. Then generate a working notebook and a dataset card documenting the deterministic splits and loading instructions for reproducibility.
Yes, dataset formatting supports temporal or deterministic partitioning with clear boundaries for train/validation/test splits. This ensures repeatable pipelines that can be version-controlled within CRISP-DM data preparation workflows.
Formatting datasets for modeling generates a working notebook at notebooks/3.5-format-data.ipynb and a structured summary document at docs/crisp-dm/3-data-preparation/3.5-format-data.md. It also produces a dataset card with loading instructions for reproducibility.
thbraetMerge cleaned datasets into a unified analysis dataset with documented join logic.
thbraetGenerate a CRISP-DM 4.2 test design with splitting strategy, metrics, and experiment tracking.
intertwinePush metadata and canonical datasets to HuggingFace repositories with gated access.
thbraetSelect datasets, fields, and records for CRISP-DM modeling datasets.
