1 | .. _Attribute Distance: |
---|

2 | |
---|

3 | Attribute Distance |
---|

4 | ================== |
---|

5 | |
---|

6 | .. image:: ../icons/AttributeDistance.png |
---|

7 | |
---|

8 | Computes distances between examples in the data set |
---|

9 | |
---|

10 | Signals |
---|

11 | ------- |
---|

12 | |
---|

13 | Inputs: |
---|

14 | - Examples |
---|

15 | A list of examples |
---|

16 | |
---|

17 | |
---|

18 | Outputs: |
---|

19 | - Distance Matrix |
---|

20 | A matrix of attribute distances |
---|

21 | |
---|

22 | |
---|

23 | Description |
---|

24 | ----------- |
---|

25 | |
---|

26 | Widget Attribute Distances computes the distances between the attributes in |
---|

27 | the data sets. Don't confuse it with a similar widget for computing the |
---|

28 | distances between examples. |
---|

29 | |
---|

30 | .. image:: images/AttributeDistance.png |
---|

31 | :alt: Association Rules Widget |
---|

32 | |
---|

33 | Since the widget cannot compute distances between discrete and continuous |
---|

34 | attributes, all attributes are first either discretized, by splitting the |
---|

35 | attribute into four quartiles, or "continuized" by treating any discrete |
---|

36 | attributes as ordinal with values equivalent to 0, 1, 2, 3... For other, |
---|

37 | possibly better methods of discretization/continuization, see widgets |
---|

38 | :ref:`Discretize` and :ref:`Continuize`. |
---|

39 | |
---|

40 | The two kinds of attributes then have different measures of distance. |
---|

41 | |
---|

42 | For discrete attributes, the distance can be computed as |
---|

43 | :obj:`Pearson's chi-square`, where the more the two attributes are dependent, |
---|

44 | the closer they are. The measure actually returns the p-value of the common |
---|

45 | chi-square test of independence. The other two measures are as defined by |
---|

46 | Aleks Jakulin in his work on `attribute interactions |
---|

47 | <http://stat.columbia.edu/~jakulin/Int/>`_: :obj:`2-way interaction` is |
---|

48 | defined as I(A;B)/H(A,B) and :obj:`3-way interaction` is I(A;B;C), |
---|

49 | respectively. |
---|

50 | |
---|

51 | |
---|

52 | Examples |
---|

53 | -------- |
---|

54 | |
---|

55 | This widget is an intermediate widget: it shows no user readable results and |
---|

56 | its output needs to be fed to a widget that can do something useful with the |
---|

57 | computed distances, for instance the :ref:`Distance Map`, |
---|

58 | :ref:`Hierarchical Clustering` to cluster the attributes, or :ref:`MDS` to |
---|

59 | visualize the distances between them. |
---|

60 | |
---|

61 | .. image:: images/AttributeDistance-Schema.png |
---|

62 | :alt: Association Rules |
---|