Skip to content

Latest commit

 

History

History
154 lines (111 loc) · 5.49 KB

Orange.data.value.rst

File metadata and controls

154 lines (111 loc) · 5.49 KB
.. py:currentmodule:: Orange.data

Values (Value)

Contains a value of a variable. Value can be discrete, continuous or of some other type, like discrete or continuous distribution, or a string.

Values are not assigned as references. For instance, when taking a value from a data instance (e.g. value = data[2]), the result is a copy of the value; changing it does not affect the original data instance.

.. attribute:: value

    The actual value.

    * If the `Value` is numeric, `value` is a number.

    * If it is discrete and the variable descriptor is present
      (:obj:`variable` is not :obj:`None`), `value` is a string with
      the symbolic value (as retrieved from
      :obj:`Orange.feature.Discrete.values`).

    * If it is discrete and the variable descriptor is unknown,
      `value` is a string with the number representing the value
      index, enclosed in brackets.

    * If the value is missing, `value` is a string ``'?'``, ``'~'``
      or ``'.'``  for don't know, don't care and other, respectively.


.. attribute:: svalue

    Stores a value that corresponds to a variable that is neither
    :obj:`Orange.feature.Discrete` nor
    :obj:`Orange.feature.Continuous` or a distribution of a
    discrete or continuous value.

    This attribute is most often used for values of
    :obj:`Orange.feature.StringVariable`; in that case `svalue`
    is an instance of :obj:`Orange.data.StringValue`. Distributions
    are seldom used; when `svalue` contains a distribution, it is
    represented with an instance of
    :obj:`Orange.statistics.distribution.Discrete` or
    :obj:`Orange.statistics.distribution.Continuous`.


.. attribute:: variable

    Descriptor related to the value. Can be ``None``.


.. attribute:: var_type

    Read-only descriptor that gives the variable type. Can be
    :obj:`Orange.feature.Type.Discrete`, :obj:`Orange.feature.Type.Continuous`,
    :obj:`Orange.feature.Type.String` or :obj:`Orange.feature.Type.Other`.


.. attribute:: value_type

    Tells whether the value is known
    (:obj:`Orange.data.Value.Regular`) or not
    (:obj:`Orange.data.Value.DC`. :obj:`Orange.data.Value.DK`).


.. method:: __init__(variable[, value])

     Construct a value with the given descriptor. Value can be any
     object that the descriptor understands. For discrete
     variables, this can be a string or an index (an integer number),
     for continuous it can be a string or a number. If the value is
     omitted, it is set to unknown (:obj:`Orange.data.Value.DK`). ::

         import Orange
         v = Orange.feature.Discrete("fruit", values=["plum", "orange", "apple"])
         an_apple = Orange.data.Value(v, "apple")
         another_apple = Orange.data.Value(v, 2)
         unknown_fruit = Orange.data.Value(v)

         v2 = Orange.feature.Continuous("iq")
         genius = Orange.data.Value(v2, 180)
         troll = Orange.data.Value(v2, "42")
         stranger = Orange.data.value(v2)

    :param variable: variable descriptor
    :type variables: Orange.feature.Descriptor
    :param value: A value
    :type value: int, float or string, or another type accepted by descriptor


.. method:: __init__(value)

    Construct either a discrete value, if the argument is an
    integer, or a continuous one, if the argument is a
    floating-point number. Descriptor is set to ``None``.

    :param value: A value
    :type value: int or float


.. method:: native()

    Return the value in a "native" Python form: strings for
    discrete and undefined values and floating-point numbers for
    continuous values.


.. method:: is_DC()

    Return ``True`` if value is "don't care".


.. method:: is_DK()

    Return ``True`` if value is "don't know".


.. method:: is_special()

    Return ``True`` if value is either "don't know" or "don't care".

Casting and comparison of values

Discrete and continuous values can be cast to Python types :obj:`int`, :obj:`float`, :obj:`long`, to strings and to boolean values. A value is considered true if it is not undefined. Continuous values support arithmetic operations.

Values can be compared with each other or with ordinary numbers. All discrete variables are treated as ordinal; values are compared by their respective indices and not in alphabetical order of their symbolic representations. When comparing values corresponding to different descriptors, Orange checks whether the order is unambiguous. Here are two such values:

deg3 = Orange.feature.Discrete(
    "deg3", values=["little", "medium", "big"])
deg4 = Orange.feature.Discrete(
    "deg4", values=["tiny", "little", "big", "huge"])
val3 = Orange.data.Value(deg3)
val4 = Orange.data.Value(deg4)
val3.value = "medium"
val4.value = "little"

Given this order, "medium" and "little" can be compared since it is known, from deg3, that "little" is less than "medium".

val3.value = "medium"
val4.value = "huge"

These two values cannot be compared since they do not appear in the same variable. (Orange cannot use transitivity to conclude that medium is less than huge since medium is less than big and big is less than huge.)

Two values also cannot be compared when they have different order in the two variables.