Phoenix Logo

phoenix_title wx.lib.plot.polyobjects.PolyBoxPlot

Creates a PolyBoxPlot object.

param data:Raw data to create a box plot from.
type data:sequence of int or float
param **attr:keyword attributes
Keyword and Default Description Type
colour='black' edge color wx.Colour
width=1 edge width float
style=wx.PENSTYLE_SOLID edge style wx.PenStyle
legend='' legend string str

Note

np.NaN and np.inf values are ignored.

TODO

  • [ ] Figure out a better way to get multiple box plots side-by-side (current method is a hack).
  • [ ] change the X axis to some labels.
  • [ ] Change getClosestPoint to only grab box plot items and outlers? Currently grabs every data point.
  • [ ] Add more customization such as Pens/Brushes, outlier shapes/size, and box width.
  • [ ] Figure out how I want to handle log-y: log data then calcBP? Or should I calc the BP first then the plot it on a log scale?

class_hierarchy Class Hierarchy

Inheritance diagram for class PolyBoxPlot:

method_summary Methods Summary

__init__  
boundingBox Returns bounding box for the plot.
calcBpData Box plot points
calcOutliers Calculates the outliers. Must be called after calcBpData.
draw Draws a box plot on the DC.
getClosestPoint Returns the index of closest point on the curve, pointXY,
getSymExtent Width and Height of Marker

api Class API



class PolyBoxPlot(PolyPoints)

Creates a PolyBoxPlot object.

Parameters:
  • data (sequence of int or float) – Raw data to create a box plot from.
  • **attr – keyword attributes
Keyword and Default Description Type
colour='black' edge color wx.Colour
width=1 edge width float
style=wx.PENSTYLE_SOLID edge style wx.PenStyle
legend='' legend string str

Note

np.NaN and np.inf values are ignored.

TODO

  • [ ] Figure out a better way to get multiple box plots side-by-side (current method is a hack).
  • [ ] change the X axis to some labels.
  • [ ] Change getClosestPoint to only grab box plot items and outlers? Currently grabs every data point.
  • [ ] Add more customization such as Pens/Brushes, outlier shapes/size, and box width.
  • [ ] Figure out how I want to handle log-y: log data then calcBP? Or should I calc the BP first then the plot it on a log scale?

Methods



__init__(self, points, **attr)


boundingBox(self)

Returns bounding box for the plot.

Override method.



calcBpData(self, data=None)

Box plot points:

Median (50%) 75% 25% low_whisker = lowest value that’s >= (25% - (IQR * 1.5)) high_whisker = highest value that’s <= 75% + (IQR * 1.5)

outliers are outside of 1.5 * IQR

data
: array-like
The data to plot
bpdata
: collections.namedtuple
Descriptive statistics for data: (min_data, low_whisker, q25, median, q75, high_whisker, max_data)


calcOutliers(self, data=None)

Calculates the outliers. Must be called after calcBpData.



wrapper(instance, dc, *args, **kwargs)

Draws a box plot on the DC.

The following draw order is required:

  1. First the whisker line
  2. Then the IQR box
  3. Lasly the median line.

This is because

  • The whiskers are drawn as single line rather than two lines
  • The median line must be visable over the box if the box has a fill.

Other than that, the draw order can be changed.



getClosestPoint(self, pntXY, pointScaled=True)

Returns the index of closest point on the curve, pointXY, scaledXY, distance x, y in user coords.

Override method.

if pointScaled == True, then based on screen coords if pointScaled == False, then based on user coords



getSymExtent(self, printerScale)

Width and Height of Marker