diff --git a/python/lsst/analysis/tools/actions/plot/histPlot.py b/python/lsst/analysis/tools/actions/plot/histPlot.py
index 2b800a929..14ff2837c 100644
--- a/python/lsst/analysis/tools/actions/plot/histPlot.py
+++ b/python/lsst/analysis/tools/actions/plot/histPlot.py
@@ -289,7 +289,7 @@ def makePlot(
             if nth_panel == 0 and nrows * ncols - len(self.panels) > 0:
                 nth_col -= 1
             # Set font size for legend based on number of panels being plotted.
-            legend_font_size = max(4, int(8 - len(self.panels[panel].hists) / 2 - nrows // 2))  # type: ignore
+            legend_font_size = max(4, int(7 - len(self.panels[panel].hists) / 2 - nrows // 2))  # type: ignore
             nums, meds, mads, stats_dict = self._makePanel(
                 data,
                 panel,
@@ -608,12 +608,16 @@ def _addStatisticsPanel(
         legend_labels = (
             ([""] * (len(handles) + 1))
             + [stats_dict["statLabels"][0]]
-            + [f"{x:.3g}" for x in stats_dict["stat1"]]
+            + [f"{x:.3g}" if x > 0.01 else f"{x:.2e}" for x in stats_dict["stat1"]]
             + [stats_dict["statLabels"][1]]
-            + [f"{x:.3g}" for x in stats_dict["stat2"]]
+            + [f"{x:.3g}" if x > 0.01 else f"{x:.2e}" for x in stats_dict["stat2"]]
             + [stats_dict["statLabels"][2]]
-            + [f"{x:.3g}" for x in stats_dict["stat3"]]
+            + [f"{x:.3g}" if x > 0.01 else f"{x:.2e}" for x in stats_dict["stat3"]]
         )
+        # Replace "e+0" with "e" and "e-0" with "e-" to save space.
+        legend_labels = [label.replace("e+0", "e") for label in legend_labels]
+        legend_labels = [label.replace("e-0", "e-") for label in legend_labels]
+
         # Set the y anchor for the legend such that it roughly lines up with
         # the panels.
         yAnchor = max(0, yAnchor0 - 0.01) + nth_col * (0.008 + len(nums) * 0.005) * legend_font_size
diff --git a/python/lsst/analysis/tools/atools/calexpMetrics.py b/python/lsst/analysis/tools/atools/calexpMetrics.py
index d1f18c16d..e92145f10 100644
--- a/python/lsst/analysis/tools/atools/calexpMetrics.py
+++ b/python/lsst/analysis/tools/atools/calexpMetrics.py
@@ -20,8 +20,15 @@
 # along with this program.  If not, see <https://www.gnu.org/licenses/>.
 from __future__ import annotations
 
-__all__ = ("CalexpSummaryMetrics",)
+__all__ = (
+    "CalexpSummaryMetrics",
+    "CalexpMetricHists",
+)
 
+from lsst.pex.config import DictField
+
+from ..actions.plot import HistPanel, HistPlot
+from ..actions.vector import BandSelector, LoadVector
 from ..interfaces import AnalysisTool
 
 
@@ -77,3 +84,24 @@ def setDefaults(self):
 
         self.prep.keysToLoad = list(self._units.keys())
         self.produce.metric.units = self._units
+
+
+class CalexpMetricHists(AnalysisTool):
+    """
+    Class to generate histograms of metrics extracted from a Metrics Table.
+    One plot per band.
+    """
+
+    parameterizedBand: bool = False
+    metrics = DictField[str, str](doc="The metrics to plot and their respective labels.")
+
+    def setDefaults(self):
+        # Band is passed as a kwarg from the calling task.
+        self.prep.selectors.bandSelector = BandSelector()
+        self.produce.plot = HistPlot()
+
+    def finalize(self):
+
+        for metric, label in self.metrics.items():
+            setattr(self.process.buildActions, metric, LoadVector(vectorKey=metric))
+            self.produce.plot.panels[metric] = HistPanel(hists={metric: "Number of calexps"}, label=label)
diff --git a/python/lsst/analysis/tools/tasks/metricAnalysis.py b/python/lsst/analysis/tools/tasks/metricAnalysis.py
index 30f14117a..fd3817507 100644
--- a/python/lsst/analysis/tools/tasks/metricAnalysis.py
+++ b/python/lsst/analysis/tools/tasks/metricAnalysis.py
@@ -26,6 +26,7 @@
 )
 
 
+from lsst.pex.config import ListField
 from lsst.pipe.base import connectionTypes as ct
 
 from ..interfaces import AnalysisBaseConfig, AnalysisBaseConnections, AnalysisPipelineTask
@@ -33,26 +34,68 @@
 
 class MetricAnalysisConnections(
     AnalysisBaseConnections,
-    dimensions=("skymap",),
-    defaultTemplates={"metricBundleName": "objectTableCore_metrics"},
+    dimensions=(),
+    defaultTemplates={"metricBundleName": ""},
 ):
+
     data = ct.Input(
-        doc="A summary table of all metrics by tract.",
+        doc="A table containing metrics.",
         name="{metricBundleName}Table",
         storageClass="ArrowAstropy",
-        dimensions=("skymap",),
         deferLoad=True,
+        dimensions=(),
     )
 
+    def __init__(self, *, config=None):
+
+        self.dimensions.update(frozenset(sorted(config.outputDataDimensions)))
+        super().__init__(config=config)
+        self.data = ct.Input(
+            doc=self.data.doc,
+            name=self.data.name,
+            storageClass=self.data.storageClass,
+            deferLoad=self.data.deferLoad,
+            dimensions=frozenset(sorted(config.inputDataDimensions)),
+        )
+
 
-class MetricAnalysisConfig(AnalysisBaseConfig, pipelineConnections=MetricAnalysisConnections):
-    pass
+class MetricAnalysisConfig(
+    AnalysisBaseConfig,
+    pipelineConnections=MetricAnalysisConnections,
+):
+    inputDataDimensions = ListField[str](
+        doc="Dimensions of the input data.",
+        default=(),
+        optional=False,
+    )
+    outputDataDimensions = ListField[str](
+        doc="Dimensions of the output data.",
+        default=(),
+        optional=False,
+    )
 
 
 class MetricAnalysisTask(AnalysisPipelineTask):
-    """Turn metric bundles which are per tract into a
-    summary metric table.
-    """
+    """Perform an analysis of a metric table."""
 
     ConfigClass = MetricAnalysisConfig
     _DefaultName = "metricAnalysis"
+
+    def runQuantum(self, butlerQC, inputRefs, outputRefs):
+
+        inputs = butlerQC.get(inputRefs)
+        dataId = butlerQC.quantum.dataId
+        plotInfo = self.parsePlotInfo(inputs, dataId)
+
+        data = self.loadData(inputs.pop("data"))
+
+        # TODO: "bands" kwarg is a workaround for DM-47941.
+        outputs = self.run(
+            data=data,
+            plotInfo=plotInfo,
+            bands=dataId["band"],
+            band=dataId["band"],
+            **inputs,
+        )
+
+        butlerQC.put(outputs, outputRefs)