Parallexaxis doc

ema_workbench/analysis/parcoords.py

@@ -143,6 +143,29 @@ class ParallelAxes:
     rot : float, optional
           rotation of axis labels
 
+    Attributes
+    ----------
+    limits : DataFrame
+             A DataFrame specifying the limits for each dimension in the
+             data set. For categorical data, the first cell should contain all
+             categories.
+    recoding : dict
+               non numeric columns are converting to integer variables
+    flipped_axes : set
+                   set of Axes that are to be shown flipped
+    axis_labels : list of str
+    fontsize : int
+    fig : a matplotlib Figure instance
+    axes : list of matplotlib Axes instances
+    ticklabels : list of str
+    datalabels : list of str
+                 labels associated with lines
+
+    Notes
+    -----
+    The basic setup of the Parallel Axis plot is a row of mpl Axes instances, with all whitespace
+    in between removed. The number of Axes is the number of columns - 1.
+
     """
 
     def __init__(self, limits, formatter=None, fontsize=14, rot=90):
@@ -155,7 +178,6 @@ def __init__(self, limits, formatter=None, fontsize=14, rot=90):
         formatter : dict, optional
                     specify precision formatters for minima and maxima,
                     defaults to .2f
-
         fontsize : int, optional
                    fontsize for defaults text items
         rot : float, optional
@@ -175,8 +197,8 @@ def __init__(self, limits, formatter=None, fontsize=14, rot=90):
                 self.recoding[column] = CategoricalDtype(categories=cats, ordered=False)
                 self.limits.loc[:, column] = [0, len(cats) - 1]
 
-        self.normalizer = preprocessing.MinMaxScaler()
-        self.normalizer.fit(self.limits)
+        self._normalizer = preprocessing.MinMaxScaler()
+        self._normalizer.fit(self.limits)
 
         fig, axes, ticklabels = setup_parallel_plot(
             self.axis_labels,
@@ -231,7 +253,7 @@ def plot(self, data, color=None, label=None, **kwargs):
             recoded[key] = data[key].astype(value).cat.codes
 
         # normalize the data
-        normalized_data = pd.DataFrame(self.normalizer.transform(recoded), columns=recoded.columns)
+        normalized_data = pd.DataFrame(self._normalizer.transform(recoded), columns=recoded.columns)
 
         # plot the data
         self._plot(normalized_data, color=color, **kwargs)

ema_workbench/examples/outputspace_exploration_lakeproblem.ipynb

@@ -475,4 +475,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}