DM-50425: Add completenessPlot publicationStyle

python/lsst/analysis/tools/actions/keyedData/calcCompletenessHistogram.py

@@ -44,7 +44,7 @@ class MagnitudeCompletenessConfig(pexConfig.Config):
     )
     completeness_percentiles = pexConfig.ListField[float](
         doc="The percentiles to find the magnitude at.",
-        default=[90.0, 80.0, 50.0, 20.0, 10.0],
+        default=[90.0, 80.0, 50.0],
         itemCheck=isPercent,
     )
 

python/lsst/analysis/tools/actions/plot/completenessPlot.py

@@ -22,11 +22,12 @@
 
 from typing import Mapping
 
-import matplotlib.pyplot as plt
 import numpy as np
 from lsst.pex.config import ChoiceField, Field
 from lsst.pex.config.configurableActions import ConfigurableActionField
-from lsst.utils.plotting import set_rubin_plotstyle
+from lsst.utils.plotting import make_figure, set_rubin_plotstyle
+
+# from matplotlib import gridspec
 from matplotlib.figure import Figure
 
 from ...actions.keyedData import CalcCompletenessHistogramAction
@@ -39,11 +40,24 @@
 class CompletenessHist(PlotAction):
     """Makes plots of completeness and purity."""
 
+    label_shift = Field[float](
+        doc="Fraction of plot width to shift completeness/purity labels by."
+        "Ignored if percentiles_style is not 'below_line'",
+        default=-0.1,
+    )
     action = ConfigurableActionField[CalcCompletenessHistogramAction](
         doc="Action to compute completeness/purity",
     )
-    mag_ref_label = Field[str](doc="Label for the completeness x axis.", default="Reference magnitude")
-    mag_target_label = Field[str](doc="Label for the purity x axis.", default="Measured magnitude")
+    color_counts = Field[str](doc="Color for the line showing object counts", default="#029E73")
+    color_right = Field[str](
+        doc="Color for the line showing the correctly classified fraction", default="#949494"
+    )
+    color_wrong = Field[str](
+        doc="Color for the line showing the wrongly classified fraction", default="#DE8F05"
+    )
+    legendLocation = Field[str](doc="Legend position within main plot", default="lower left")
+    mag_ref_label = Field[str](doc="Label for the completeness x axis.", default="Reference Magnitude")
+    mag_target_label = Field[str](doc="Label for the purity x axis.", default="Measured Magnitude")
     percentiles_style = ChoiceField[str](
         doc="Style and locations for completeness threshold percentile labels",
         allowed={
@@ -52,6 +66,8 @@ class CompletenessHist(PlotAction):
         },
         default="below_line",
     )
+    publicationStyle = Field[bool](doc="Make a publication-style of plot", default=False)
+    show_purity = Field[bool](doc="Whether to include a purity plot below completness", default=True)
 
     def getInputSchema(self) -> KeyedDataSchema:
         yield from self.action.getOutputSchema()
@@ -132,10 +148,12 @@ def makePlot(self, data, plotInfo, **kwargs):
 
         # Make plot showing the fraction recovered in magnitude bins
         set_rubin_plotstyle()
-        fig, axes = plt.subplots(dpi=300, nrows=2, figsize=(8, 8))
-        color_counts = "purple"
-        color_wrong = "firebrick"
-        color_right = "teal"
+        n_sub = 1 + self.show_purity
+        fig = make_figure(dpi=300, figsize=(8, 4 * n_sub))
+        if self.show_purity:
+            axes = (fig.add_subplot(2, 1, 1), fig.add_subplot(2, 1, 2))
+        else:
+            axes = [fig.add_axes([0.1, 0.15, 0.8, 0.75])]
         max_left = 1.05
 
         band = kwargs.get("band")
@@ -167,28 +185,37 @@ def makePlot(self, data, plotInfo, **kwargs):
 
         counts_all = data[names["count"]]
 
+        if self.publicationStyle:
+            lineTuples = (
+                (data[names["completeness"]], False, "k", "Completeness"),
+                (data[names["completeness_bad_match"]], False, self.color_wrong, "Incorrect Class"),
+            )
+        else:
+            lineTuples = (
+                (data[names["completeness"]], True, "k", "Completeness"),
+                (data[names["completeness_bad_match"]], False, self.color_wrong, "Incorrect class"),
+                (data[names["completeness_good_match"]], False, self.color_right, "Correct Class"),
+            )
+
         plots = {
             "Completeness": {
                 "count_type": "Reference",
                 "counts": data[names["count_ref"]],
-                "lines": (
-                    (data[names["completeness"]], True, "k", "completeness"),
-                    (data[names["completeness_bad_match"]], False, color_wrong, "wrong class"),
-                    (data[names["completeness_good_match"]], False, color_right, "right class"),
-                ),
+                "lines": lineTuples,
                 "xlabel": self.mag_ref_label,
             },
-            "Purity": {
+        }
+        if self.show_purity:
+            plots["Purity"] = {
                 "count_type": "Object",
                 "counts": data[names["count_target"]],
                 "lines": (
-                    (data[names["purity"]], True, "k", None),
-                    (data[names["purity_bad_match"]], False, color_wrong, "wrong class"),
-                    (data[names["purity_good_match"]], False, color_right, "right class"),
+                    (data[names["purity"]], True, "k", "Purity"),
+                    (data[names["purity_bad_match"]], False, self.color_wrong, "Incorrect class"),
+                    (data[names["purity_good_match"]], False, self.color_right, "Correct class"),
                 ),
                 "xlabel": self.mag_target_label,
-            },
-        }
+            }
 
         # idx == 0 should be completeness; update this if that assumption
         # is changed
@@ -203,9 +230,10 @@ def makePlot(self, data, plotInfo, **kwargs):
                 xticks=np.arange(round(xlim[0]), round(xlim[1])),
                 yticks=np.linspace(0, 1, 11),
             )
-            axes_idx.grid(color="lightgrey", ls="-")
+            if not self.publicationStyle:
+                axes_idx.grid(color="lightgrey", ls="-")
             ax_right = axes_idx.twinx()
-            ax_right.set_ylabel(f"{plot_data['count_type']} counts/mag")
+            ax_right.set_ylabel(f"{plot_data['count_type']} Counts/Magnitude", color="k")
             ax_right.set_yscale("log")
 
             for y, do_err, color, label in plot_data["lines"]:
@@ -214,26 +242,43 @@ def makePlot(self, data, plotInfo, **kwargs):
                     y=y,
                     xerr=x_err if do_err else None,
                     yerr=1.0 / np.sqrt(counts_all + 1) if do_err else None,
+                    capsize=0,
                     color=color,
                     label=label,
                 )
             y = plot_data["counts"] / interval
             # It should be unusual for np.max(y) to be zero; nonetheless...
+            lines_left, labels_left = axes_idx.get_legend_handles_labels()
             ax_right.step(
                 [x[0] - interval] + list(x) + [x[-1] + interval],
                 [0] + list(y) + [0],
                 where="mid",
-                color=color_counts,
-                label="counts",
+                color=self.color_counts,
+                label="Counts",
             )
+
+            # Force the inputs counts histogram to the back
+            ax_right.zorder = 1
+            axes_idx.zorder = 2
+            axes_idx.patch.set_visible(False)
+
             ax_right.set_ylim(0.999, 10 ** (max_left * np.log10(max(np.nanmax(y), 2))))
-            ax_right.tick_params(axis="y", labelcolor=color_counts)
-            lines_left, labels_left = axes_idx.get_legend_handles_labels()
+            ax_right.tick_params(axis="y", labelcolor=self.color_counts)
             lines_right, labels_right = ax_right.get_legend_handles_labels()
-            axes_idx.legend(lines_left + lines_right, labels_left + labels_right, loc="lower left", ncol=2)
+
+            # Using fig for legend
+            (axes_idx if self.show_purity else fig).legend(
+                lines_left + lines_right,
+                labels_left + labels_right,
+                loc=self.legendLocation,
+                ncol=2,
+            )
 
             if idx == 0:
-                percentiles = self.action.config_metrics.completeness_percentiles
+                if not self.publicationStyle:
+                    percentiles = self.action.config_metrics.completeness_percentiles
+                else:
+                    percentiles = [90.0, 50.0]
                 if percentiles:
                     above_plot = self.percentiles_style == "above_plot"
                     below_line = self.percentiles_style == "below_line"
@@ -242,7 +287,7 @@ def makePlot(self, data, plotInfo, **kwargs):
                     if above_plot:
                         texts = []
                     elif below_line:
-                        offset = 0.1 * (xlims[1] - xlims[0])
+                        offset = self.label_shift * (xlims[1] - xlims[0])
                     else:
                         raise RuntimeError(f"Unimplemented {self.percentiles_style=}")
                     for pct in percentiles:
@@ -260,13 +305,22 @@ def makePlot(self, data, plotInfo, **kwargs):
                             if above_plot:
                                 texts.append(text)
                             elif below_line:
-                                axes_idx.text(mag_completeness - offset, pct, text, ha="right", va="top")
+                                axes_idx.text(
+                                    mag_completeness + offset,
+                                    pct - 0.02,
+                                    text,
+                                    ha="right",
+                                    va="top",
+                                    fontsize=12,
+                                )
                     if above_plot:
                         texts = f"Thresholds: {'; '.join(texts)}"
                         axes_idx.text(xlims[0], max_left, texts, ha="left", va="bottom")
 
         # Add useful information to the plot
-        addPlotInfo(fig, plotInfo)
-        fig.tight_layout()
-        fig.subplots_adjust(top=0.90)
+        if not self.publicationStyle:
+            addPlotInfo(fig, plotInfo)
+        if self.show_purity:
+            fig.tight_layout()
+            fig.subplots_adjust(top=0.90)
         return fig

python/lsst/analysis/tools/actions/plot/plotUtils.py

@@ -20,7 +20,7 @@
 # along with this program.  If not, see <https://www.gnu.org/licenses/>.
 from __future__ import annotations
 
-__all__ = ("PanelConfig",)
+__all__ = ("PanelConfig", "sortAllArrays")
 
 from typing import TYPE_CHECKING, Iterable, List, Mapping, Tuple