datasets

---

iso_doc_type: "Specification" iso_viewpoint: "ComponentView" type: "module" title: "regression_model_template::io::datasets — Tabular Dataset I/O & Lineage" source_path: "src/regression_model_template/io/datasets.py" description: "Abstract Reader/Writer interfaces and concrete Parquet implementations with automated lineage extraction." tags: ["iso42010", "okf", "component_view", "datasets", "parquet", "lineage"] timestamp: "2026-10-01T17:50:00Z" generated: "agent:wiki-iso-documentation-standard" verified: "true" last_verified_commit: "073b5fd"

regression_model_template::io::datasets — Tabular Dataset I/O & Lineage

Source: src/regression_model_template/io/datasets.py (Lines: L1-L128)
Layer: Infrastructure
Role: Standardized reading and writing of tabular Parquet datasets with automated data lineage recording for MLflow tracking.

---

1. Class Diagram

classDiagram
    class Reader {
        <<abstract>>
        +str KIND
        +int limit
        +read() DataFrame*
        +lineage(name: str, data: DataFrame, targets, predictions) Lineage
    }

    class ParquetReader {
        +str KIND = "ParquetReader"
        +str path
        +read() DataFrame
        +lineage(name: str, data: DataFrame, targets, predictions) Lineage
    }
    ParquetReader --|> Reader : implements

    class Writer {
        <<abstract>>
        +str KIND
        +write(data: DataFrame) None*
    }

    class ParquetWriter {
        +str KIND = "ParquetWriter"
        +str path
        +write(data: DataFrame) None
    }
    ParquetWriter --|> Writer : implements

Loading

---

Related: Schemas · Data Model