Layers

Layers are composable transformation stages that an entity flows through. Each layer takes the output of the previous layer and applies additional transformations.

Available Layers

Layer	Purpose
PrepLayer	Clean raw data: type casts, renames, computed columns
DimensionLayer	Add business logic columns (is_paying, account_type)
SnapshotLayer	Track changes over time (daily snapshots, SCD)
ActivityConfig	Detect events from state changes (row_appears, status_becomes, field_changes)
AnalyticsLayer	Date-grain metric aggregation (snapshot and event metrics)

Layer Configuration

All layers are configured inside LayersConfig on the entity:

from fyrnheim import (
    LayersConfig, PrepLayer, DimensionLayer, SnapshotLayer,
    ActivityConfig, ActivityType, AnalyticsLayer, AnalyticsMetric,
    ComputedColumn,
)

layers = LayersConfig(
    prep=PrepLayer(
        model_name="prep_customers",
        computed_columns=[
            ComputedColumn(name="amount_dollars", expression="t.amount_cents / 100.0"),
        ],
    ),
    dimension=DimensionLayer(
        model_name="dim_customers",
        computed_columns=[
            ComputedColumn(name="is_paying", expression="t.plan != 'free'"),
        ],
    ),
    activity=ActivityConfig(
        model_name="activity_customers",
        entity_id_field="customer_id",
        types=[
            ActivityType(
                name="signed_up",
                trigger="row_appears",
                timestamp_field="created_at",
            ),
        ],
    ),
    analytics=AnalyticsLayer(
        model_name="analytics_customers",
        date_expression="t.created_at.date()",
        metrics=[
            AnalyticsMetric(
                name="new_customers",
                expression="t.count()",
                metric_type="event",
            ),
        ],
    ),
)

PrepLayer

The first transformation stage. Use it to clean raw data with type casts, renames, and computed columns.

PrepLayer(
    model_name="prep_customers",
    computed_columns=[
        ComputedColumn(name="email_hash", expression=hash_email("email")),
        ComputedColumn(name="amount_dollars", expression="t.amount_cents / 100.0"),
    ],
)

DimensionLayer

Add business logic columns derived from prep layer output.

DimensionLayer(
    model_name="dim_customers",
    computed_columns=[
        ComputedColumn(name="is_paying", expression="t.plan != 'free'"),
        ComputedColumn(name="account_type", expression=categorize("t.plan", {"free": "free", "pro": "paid", "enterprise": "paid"})),
    ],
)

SnapshotLayer

Track changes over time with daily snapshots, useful for slowly changing dimensions (SCD).

ActivityConfig

Detect events from state changes. Supports triggers like row_appears, status_becomes, and field_changes.

ActivityConfig(
    model_name="activity_customers",
    entity_id_field="customer_id",
    types=[
        ActivityType(name="signed_up", trigger="row_appears", timestamp_field="created_at"),
    ],
)

AnalyticsLayer

Date-grain metric aggregation for both snapshot and event metrics.

AnalyticsLayer(
    model_name="analytics_customers",
    date_expression="t.created_at.date()",
    metrics=[
        AnalyticsMetric(name="new_customers", expression="t.count()", metric_type="event"),
    ],
)