feat!: Introduce new summary section for data metrics

diffly/cli.py

@@ -12,18 +12,11 @@
 
 from ._compat import typer
 from ._utils import ABS_TOL_DEFAULT, ABS_TOL_TEMPORAL_DEFAULT, REL_TOL_DEFAULT
-from .metrics import Metric, MetricFn
 from .metrics.change import DEFAULT_CHANGE_METRICS
 from .metrics.data import DEFAULT_DATA_METRICS
 
 app = typer.Typer()
 
-#: All metric presets selectable via ``--metric``, combining the change and data sets.
-AVAILABLE_METRICS: dict[str, MetricFn | Metric] = {
-    **DEFAULT_CHANGE_METRICS,
-    **DEFAULT_DATA_METRICS,
-}
-
 
 @app.command()
 def main(
@@ -143,12 +136,21 @@ def main(
         list[str],
         typer.Option(hidden=True),
     ] = [],
-    metric: Annotated[
+    data_metric: Annotated[
         list[str],
         typer.Option(
             help=(
-                "Metric presets to display per column. Repeatable. "
-                f"Available: {', '.join(AVAILABLE_METRICS)}."
+                "Data metric presets to display in the Data Inspection section. "
+                f"Repeatable. Available: {', '.join(DEFAULT_DATA_METRICS)}."
+            )
+        ),
+    ] = [],
+    change_metric: Annotated[
+        list[str],
+        typer.Option(
+            help=(
+                "Change metric presets to display as extra columns in the Columns "
+                f"table. Repeatable. Available: {', '.join(DEFAULT_CHANGE_METRICS)}."
             )
         ),
     ] = [],
@@ -161,12 +163,20 @@ def main(
         )
         hidden_column = [*hidden_column, *hidden_columns]
 
-    for name in metric:
-        if name not in AVAILABLE_METRICS:
+    for name in data_metric:
+        if name not in DEFAULT_DATA_METRICS:
+            raise typer.BadParameter(
+                f"Unknown data metric: {name!r}. "
+                f"Available: {', '.join(DEFAULT_DATA_METRICS)}."
+            )
+    for name in change_metric:
+        if name not in DEFAULT_CHANGE_METRICS:
             raise typer.BadParameter(
-                f"Unknown metric: {name!r}. Available: {', '.join(AVAILABLE_METRICS)}."
+                f"Unknown change metric: {name!r}. "
+                f"Available: {', '.join(DEFAULT_CHANGE_METRICS)}."
             )
-    metrics = {name: AVAILABLE_METRICS[name] for name in metric}
+    data_metrics = {name: DEFAULT_DATA_METRICS[name] for name in data_metric}
+    change_metrics = {name: DEFAULT_CHANGE_METRICS[name] for name in change_metric}
 
     comparison = compare_frames(
         pl.scan_parquet(left),
@@ -185,7 +195,8 @@ def main(
         right_name=right_name,
         slim=slim,
         hidden_columns=hidden_column,
-        metrics=metrics,
+        data_metrics=data_metrics,
+        change_metrics=change_metrics,
     )
     if output_json:
         typer.echo(summary.to_json())

diffly/comparison.py

@@ -25,7 +25,8 @@
     lazy_len,
     make_and_validate_mapping,
 )
-from .metrics import Metric, MetricFn, _make_numeric_metric
+from .metrics.change import ChangeMetric, ChangeMetricFn
+from .metrics.data import DataMetric, DataMetricFn
 
 if TYPE_CHECKING:  # pragma: no cover
     # NOTE: We cannot import at runtime as we're otherwise running into circular
@@ -920,7 +921,8 @@ def summary(
         right_name: str = Side.RIGHT,
         slim: bool = False,
         hidden_columns: list[str] | None = None,
-        metrics: Mapping[str, MetricFn | Metric] | None = None,
+        data_metrics: Mapping[str, DataMetricFn | DataMetric] | None = None,
+        change_metrics: Mapping[str, ChangeMetricFn | ChangeMetric] | None = None,
     ) -> Summary:
         """Generate a summary of all aspects of the comparison.
 
@@ -950,18 +952,26 @@ def summary(
                 advanced users who are familiar with the summary format.
             hidden_columns: Columns for which no values are printed, e.g. because they
                 contain sensitive information.
-            metrics: Optional mapping from display label to a metric. A value may be a
-                callable ``(left_expr, right_expr) -> pl.Expr`` or a
-                :class:`~diffly.metrics.Metric`. Each callable receives two
-                :class:`polars.Expr` referring to the left and right values of a single
-                column across all joined rows, and must return a scalar aggregation
-                expression. Bare callables are only computed for numerical columns; wrap
-                one in a :class:`~diffly.metrics.Metric` with a column selector to target
-                other column types (e.g. ``Metric(fn, selector=cs.all())``).
-                See :doc:`/api/metrics` for the full list of presets and the
-                :data:`~diffly.metrics.MetricFn` type. When ``None`` (default), no metrics
-                are computed; presets are not applied automatically. Prefer short labels —
-                the summary has a fixed width and many or long labels degrade rendering.
+            data_metrics: Optional mapping from display label to a data metric,
+                describing each dataset individually and rendered in the "Data
+                Inspection" section. A value may be a
+                :class:`~diffly.metrics.data.DataMetric` or a bare callable taking a
+                single column expression, which is wrapped in a
+                :class:`~diffly.metrics.data.DataMetric` applying to all columns. To
+                target other column types, construct the metric explicitly with a
+                column selector (e.g. ``DataMetric(fn, selector=cs.numeric())``).
+            change_metrics: Optional mapping from display label to a change metric,
+                quantifying the change between the two sides and rendered as extra
+                columns in the "Columns" table. A value may be a
+                :class:`~diffly.metrics.change.ChangeMetric` or a bare callable taking a
+                pair of column expressions, which is wrapped in a
+                :class:`~diffly.metrics.change.ChangeMetric` applying to numerical
+                columns. To target other column types, construct the metric explicitly
+                with a column selector (e.g. ``ChangeMetric(fn, selector=cs.numeric())``).
+                See :doc:`/api/metrics` for the full list of presets. When ``None``
+                (default), no metrics are computed; presets are not applied
+                automatically. Prefer short labels — the summary has a fixed width and
+                many or long labels degrade rendering.
 
         Returns:
             A summary which can be printed or written to a file.
@@ -977,12 +987,14 @@ def summary(
         # NOTE: We're importing here to prevent circular imports
         from .summary import Summary
 
-        resolved_metrics = (
-            {
-                label: v if isinstance(v, Metric) else _make_numeric_metric(v)
-                for label, v in metrics.items()
-            }
-            if metrics is not None
+        resolved_data_metrics = (
+            {label: _resolve_data_metric(v) for label, v in data_metrics.items()}
+            if data_metrics is not None
+            else None
+        )
+        resolved_change_metrics = (
+            {label: _resolve_change_metric(v) for label, v in change_metrics.items()}
+            if change_metrics is not None
             else None
         )
 
@@ -996,7 +1008,8 @@ def summary(
             right_name=right_name,
             slim=slim,
             hidden_columns=hidden_columns,
-            metrics=resolved_metrics,
+            data_metrics=resolved_data_metrics,
+            change_metrics=resolved_change_metrics,
         )
 
     # ----------------------------------- UTILITIES ----------------------------------- #
@@ -1239,3 +1252,11 @@ def _list_length_exprs(
             for e in _list_length_exprs(expr.struct[field.name], field.dtype)
         ]
     return []
+
+
+def _resolve_data_metric(v: DataMetricFn | DataMetric) -> DataMetric:
+    return v if isinstance(v, DataMetric) else DataMetric(fn=v)
+
+
+def _resolve_change_metric(v: ChangeMetricFn | ChangeMetric) -> ChangeMetric:
+    return v if isinstance(v, ChangeMetric) else ChangeMetric(fn=v)

diffly/metrics/__init__.py

@@ -1,50 +1,17 @@
 # Copyright (c) QuantCo 2025-2026
 # SPDX-License-Identifier: BSD-3-Clause
-
 """Metrics computed per column when generating a summary.
 
 Two families are provided:
 
-- Metrics in :mod:`~diffly.metrics.change` describe the change between numeric
-  columns itself by aggregating over ``right - left``.
-- Metrics in :mod:`~diffly.metrics.data` describe the left and right datasets
-  individually, explaining how a change affects the data.
+- :class:`~diffly.metrics.change.ChangeMetric`s in :mod:`~diffly.metrics.change` describe the change between
+  columns itself by aggregating over a combination of the columns (e.g., ``right - left``).
+- :class:`~diffly.metrics.data.DataMetric`s in :mod:`~diffly.metrics.data` describe the left and right
+  datasets individually, explaining how a change affects the data.
 """
 
 from __future__ import annotations
 
 from . import change, data
-from ._common import Metric, MetricFn
-from .change import (
-    _make_numeric_metric,
-    max,
-    mean,
-    mean_absolute_deviation,
-    mean_relative_deviation,
-    median,
-    min,
-    quantile,
-    std,
-)
-
-DEFAULT_METRICS: dict[str, MetricFn | Metric] = {
-    **change.DEFAULT_CHANGE_METRICS,
-}
-"""The default preset metrics, consisting of the change default set."""
 
-__all__ = [
-    "DEFAULT_METRICS",
-    "Metric",
-    "MetricFn",
-    "change",
-    "data",
-    "max",
-    "mean",
-    "mean_absolute_deviation",
-    "mean_relative_deviation",
-    "median",
-    "min",
-    "quantile",
-    "std",
-    "_make_numeric_metric",
-]
+__all__ = ["change", "data"]

diffly/metrics/change.py

@@ -1,21 +1,37 @@
 # Copyright (c) QuantCo 2025-2026
 # SPDX-License-Identifier: BSD-3-Clause
 
-"""Metrics describing the change between numeric columns.
-
-These aggregate over ``right - left`` to characterize the change itself.
-"""
-
 from __future__ import annotations
 
+from collections.abc import Callable
+from dataclasses import dataclass, field
+
 import polars as pl
 import polars.selectors as cs
 
-from ._common import Metric, MetricFn
+ChangeMetricFn = Callable[[pl.Expr, pl.Expr], pl.Expr]
+"""A `ChangeMetricFn` maps a pair of column expressions to a scalar aggregation
+expression."""
+
+
+@dataclass(frozen=True)
+class ChangeMetric:
+    """A metric quantifying the *change* in a column between the two sides of a
+    comparison.
+
+    Change metrics are rendered as extra columns in the "Columns" table, alongside the
+    match rate.
+    """
+
+    fn: ChangeMetricFn
+    """Aggregates over ``right - left`` (e.g. the mean delta) to describe the change
+    itself."""
+
+    selector: cs.Selector = field(default_factory=cs.numeric)
+    """Selects the columns the metric applies to; defaults to numeric columns."""
 
 
-def _make_numeric_metric(fn: MetricFn) -> Metric:
-    return Metric(fn=fn, selector=cs.numeric())
+# ---------------------------------- CHANGE METRICS ---------------------------------- #
 
 
 def mean(left: pl.Expr, right: pl.Expr) -> pl.Expr:
@@ -54,7 +70,7 @@ def mean_relative_deviation(left: pl.Expr, right: pl.Expr) -> pl.Expr:
     return ((right - left) / left).abs().mean()
 
 
-def quantile(q: float) -> MetricFn:
+def quantile(q: float) -> ChangeMetricFn:
     """Factory returning a metric that computes the ``q``-quantile of
     ``right - left``."""
     if not 0 <= q <= 1:
@@ -66,13 +82,13 @@ def _quantile(left: pl.Expr, right: pl.Expr) -> pl.Expr:
     return _quantile
 
 
-DEFAULT_CHANGE_METRICS: dict[str, MetricFn] = {
-    "Mean": mean,
-    "Median": median,
-    "Min": min,
-    "Max": max,
-    "Std": std,
-    "Mean absolute deviation": mean_absolute_deviation,
-    "Mean relative deviation": mean_relative_deviation,
+DEFAULT_CHANGE_METRICS: dict[str, ChangeMetric] = {
+    "Mean diff": ChangeMetric(fn=mean),
+    "Median diff": ChangeMetric(fn=median),
+    "Min diff": ChangeMetric(fn=min),
+    "Max diff": ChangeMetric(fn=max),
+    "Std diff": ChangeMetric(fn=std),
+    "Mean absolute diff": ChangeMetric(fn=mean_absolute_deviation),
+    "Mean relative diff": ChangeMetric(fn=mean_relative_deviation),
 }
 """Preset metrics describing the change between numeric columns."""