inspect multiple boxes and display them in a single figure (#317)

* make it possible to plot mulitple boxes in one figure

default behavior of inspect with list of indices is to show each box in a seperate figure (if style is graph). This code makes it possible to plot them in a single figure.

closes #124

* Update test_prim.py

ema_workbench/analysis/prim.py

@@ -19,6 +19,7 @@
 from operator import itemgetter
 
 import matplotlib as mpl
+import matplotlib.pyplot as plt
 import numpy as np
 import pandas as pd
 import seaborn as sns
@@ -396,9 +397,9 @@ def __getattr__(self, name):
         else:
             raise AttributeError
 
-    def inspect(self, i=None, style="table", **kwargs):
+    def inspect(self, i=None, style="table", ax=None, **kwargs):
         """Write the stats and box limits of the user specified box to
-        standard out. if i is not provided, the last box will be
+        standard out. If i is not provided, the last box will be
         printed
 
         Parameters
@@ -409,20 +410,38 @@ def inspect(self, i=None, style="table", **kwargs):
                 the style of the visualization. 'table' prints the stats and
                 boxlim. 'graph' creates a figure. 'data' returns a list of
                 tuples, where each tuple contains the stats and the box_lims.
+        ax : axes or list of axes instances, optional
+             used in conjunction with `graph` style, allows you to control the axes on which graph is plotted
+             if i is list, axes should be list of equal length. If axes is None, each i_j in i will be plotted
+             in a separate figure.
 
         additional kwargs are passed to the helper function that
         generates the table or graph
 
         """
+        if style not in {"table", "graph", "data"}:
+            raise ValueError(f"style must be one of 'table', 'graph', or 'data', not {style}")
+
         if i is None:
             i = [self._cur_box]
         elif isinstance(i, int):
             i = [i]
 
+        if isinstance(ax, mpl.axes.Axes):
+            ax = [ax]
+
         if not all(isinstance(x, int) for x in i):
             raise TypeError(f"i must be an integer or list of integers, not {type(i)}")
 
-        return [self._inspect(entry, style=style, **kwargs) for entry in i]
+        if (ax is not None) and style == "graph":
+            if len(ax) != len(i):
+                raise ValueError(
+                    f"the number of axes ({len(ax)}) does not match the number of boxes to inspect ({len(i)})"
+                )
+            else:
+                return [self._inspect(i_j, style=style, ax=ax, **kwargs) for i_j, ax in zip(i, ax)]
+        else:
+            return [self._inspect(entry, style=style, **kwargs) for entry in i]
 
     def _inspect(self, i=None, style="table", **kwargs):
         """Helper method for inspecting one or more boxes on the
@@ -450,7 +469,13 @@ def _inspect(self, i=None, style="table", **kwargs):
         if style == "table":
             return self._inspect_table(i, uncs, qp_values)
         elif style == "graph":
-            return self._inspect_graph(i, uncs, qp_values, **kwargs)
+            # makes it possible to use _inspect to plot multiple
+            # boxes into a single figure
+            try:
+                ax = kwargs.pop("ax")
+            except KeyError:
+                fig, ax = plt.subplots()
+            return self._inspect_graph(i, uncs, qp_values, ax=ax, **kwargs)
         elif style == "data":
             return self._inspect_data(i, uncs, qp_values)
         else:
@@ -496,6 +521,7 @@ def _inspect_graph(
         ticklabel_formatter="{} ({})",
         boxlim_formatter="{: .2g}",
         table_formatter="{:.3g}",
+        ax=None,
     ):
         """Helper method for visualizing box statistics in
         graph form"""
@@ -507,6 +533,7 @@ def _inspect_graph(
             uncs,
             self.peeling_trajectory.at[i, "coverage"],
             self.peeling_trajectory.at[i, "density"],
+            ax,
             ticklabel_formatter=ticklabel_formatter,
             boxlim_formatter=boxlim_formatter,
             table_formatter=table_formatter,

ema_workbench/analysis/scenario_discovery_util.py

@@ -324,6 +324,7 @@ def plot_pair_wise_scatter(
     fill_subplots=True,
 ):
     """helper function for pair wise scatter plotting
+
     Parameters
     ----------
     x : DataFrame
@@ -530,16 +531,19 @@ def plot_pair_wise_scatter(
     return grid
 
 
-def _setup_figure(uncs):
+def _setup_figure(uncs, ax):
     """
 
     helper function for creating the basic layout for the figures that
     show the box lims.
 
+    Parameters
+    ----------
+    uncs : list of str
+    ax : axes instance
+
     """
     nr_unc = len(uncs)
-    fig = plt.figure()
-    ax = fig.add_subplot(111)
 
     # create the shaded grey background
     rect = mpl.patches.Rectangle(
@@ -551,7 +555,6 @@ def _setup_figure(uncs):
     ax.yaxis.set_ticks(list(range(nr_unc)))
     ax.xaxis.set_ticks([0, 0.25, 0.5, 0.75, 1])
     ax.set_yticklabels(uncs[::-1])
-    return fig, ax
 
 
 def plot_box(
@@ -561,6 +564,7 @@ def plot_box(
     uncs,
     coverage,
     density,
+    ax,
     ticklabel_formatter="{} ({})",
     boxlim_formatter="{: .2g}",
     table_formatter="{:.3g}",
@@ -579,6 +583,7 @@ def plot_box(
     ticklabel_formatter : str
     boxlim_formatter : str
     table_formatter : str
+    ax : Axes instance
 
     Returns
     -------
@@ -587,8 +592,9 @@ def plot_box(
 
     """
     norm_box_lim = _normalize(boxlim, box_init, uncs)
+    fig = plt.gcf()
 
-    fig, ax = _setup_figure(uncs)
+    _setup_figure(uncs, ax)
     for j, u in enumerate(uncs):
         # we want to have the most restricted dimension
         # at the top of the figure
@@ -842,7 +848,8 @@ def plot_boxes(x, boxes, together):
     norm_box_lims = [_normalize(box_lim, box_init, uncs) for box_lim in boxes]
 
     if together:
-        fig, ax = _setup_figure(uncs)
+        fig, ax = plt.subplots()
+        _setup_figure(uncs, ax)
 
         for i, u in enumerate(uncs):
             colors = itertools.cycle(COLOR_LIST)
@@ -862,7 +869,8 @@ def plot_boxes(x, boxes, together):
         colors = itertools.cycle(COLOR_LIST)
 
         for j, norm_box_lim in enumerate(norm_box_lims):
-            fig, ax = _setup_figure(uncs)
+            fig, ax = plt.subplots()
+            _setup_figure(uncs, ax)
             ax.set_title(f"box {j}")
             color = next(colors)
 

ema_workbench/examples/sd_prim_flu.py

@@ -35,7 +35,13 @@ def classify(data):
 box_1 = prim_obj.find_box()
 box_1.show_ppt()
 box_1.show_tradeoff()
-box_1.inspect(5, style="graph", boxlim_formatter="{: .2f}")
+# box_1.inspect([5, 6], style="graph", boxlim_formatter="{: .2f}")
+
+fig, axes = plt.subplots(nrows=2, ncols=1)
+
+box_1.inspect([5, 6], style="graph", boxlim_formatter="{: .2f}", ax=axes)
+plt.show()
+
 box_1.inspect(5)
 box_1.select(5)
 box_1.write_ppt_to_stdout()

test/test_analysis/test_prim.py

@@ -8,6 +8,7 @@
 
 import numpy as np
 import pandas as pd
+import matplotlib.pyplot as plt
 
 from ema_workbench.analysis import prim
 from ema_workbench.analysis.prim import PrimBox
@@ -69,9 +70,19 @@ def test_inspect(self):
         box.inspect(1)
         box.inspect()
         box.inspect(style="graph")
+        box.inspect(style="data")
 
         box.inspect([0, 1])
 
+        fig, axes = plt.subplots(2)
+        box.inspect([0, 1], ax=axes, style="graph")
+
+        fig, ax = plt.subplots()
+        box.inspect(0, ax=ax, style="graph")
+
+        with pytest.raises(ValueError):
+            fig, axes = plt.subplots(3)
+            box.inspect([0, 1], ax=axes, style="graph")
         with pytest.raises(ValueError):
             box.inspect(style="some unknown style")
         with pytest.raises(TypeError):

test/test_analysis/test_scenario_discovery_util.py

@@ -185,7 +185,8 @@ def test_plot_box(self):
 
         qp_values = {"a": [0.05, 0.9], "c": [0.05, -1]}
 
-        sdutil.plot_box(boxlim, qp_values, box_init, restricted_dims, 1, 1)
+        fig, ax = plt.subplots()
+        sdutil.plot_box(boxlim, qp_values, box_init, restricted_dims, 1, 1, ax)
         plt.draw()
         plt.close("all")