Changeset 8010:38e6a5a6eabc in orange


Ignore:
Timestamp:
06/14/11 13:00:19 (3 years ago)
Author:
markotoplak
Branch:
default
Convert:
5aa62045b82a0767461553ef95d21e9773d85121
Message:

Documentation reorganization: title needs to be in the file, not in the place where is is linked from.

Location:
orange
Files:
9 edited

Legend:

Unmodified
Added
Removed
  • orange/Orange/misc/__init__.py

    r8008 r8010  
    44 
    55Module Orange.misc contains common functions and classes which are used in other modules. 
    6  
    7 ================== 
    8 Counters 
    9 ================== 
    10  
    11 .. index:: misc 
    12 .. index:: 
    13    single: misc; counters 
    146 
    157.. automodule:: Orange.misc.counters 
    168  :members: 
    179 
    18 ================== 
    19 Render 
    20 ================== 
    21  
    22 .. index:: misc 
    23 .. index:: 
    24    single: misc; render 
    25  
    2610.. automodule:: Orange.misc.render 
    2711  :members: 
    2812 
    29 ================== 
    30 Selection 
    31 ================== 
    32  
    33 .. index:: selection 
    34 .. index:: 
    35    single: misc; selection 
    36  
    37 Many machine learning techniques generate a set different solutions or have to 
    38 choose, as for instance in classification tree induction, between different 
    39 features. The most trivial solution is to iterate through the candidates, 
    40 compare them and remember the optimal one. The problem occurs, however, when 
    41 there are multiple candidates that are equally good, and the naive approaches 
    42 would select the first or the last one, depending upon the formulation of 
    43 the if-statement. 
    44  
    45 :class:`Orange.misc.selection` provides a class that makes a random choice 
    46 in such cases. Each new candidate is compared with the currently optimal 
    47 one; it replaces the optimal if it is better, while if they are equal, 
    48 one is chosen by random. The number of competing optimal candidates is stored, 
    49 so in this random choice the probability to select the new candidate (over the 
    50 current one) is 1/w, where w is the current number of equal candidates, 
    51 including the present one. One can easily verify that this gives equal 
    52 chances to all candidates, independent of the order in which they are presented. 
    53  
    5413.. automodule:: Orange.misc.selection 
    5514  :members: 
    5615 
    57 Example 
    58 -------- 
    59  
    60 The following snippet loads the data set lymphography and prints out the 
    61 feature with the highest information gain. 
    62  
    63 part of `misc-selection-bestonthefly.py`_ (uses `lymphography.tab`_) 
    64  
    65 .. literalinclude:: code/misc-selection-bestonthefly.py 
    66   :lines: 7-16 
    67  
    68 Our candidates are tuples gain ratios and features, so we set 
    69 :obj:`callCompareOn1st` to make the compare function compare the first element 
    70 (gain ratios). We could achieve the same by initializing the object like this: 
    71  
    72 part of `misc-selection-bestonthefly.py`_ (uses `lymphography.tab`_) 
    73  
    74 .. literalinclude:: code/misc-selection-bestonthefly.py 
    75   :lines: 18-18 
    76  
    77  
    78 The other way to do it is through indices. 
    79  
    80 `misc-selection-bestonthefly.py`_ (uses `lymphography.tab`_) 
    81  
    82 .. literalinclude:: code/misc-selection-bestonthefly.py 
    83   :lines: 25- 
    84  
    85 .. _misc-selection-bestonthefly.py: code/misc-selection-bestonthefly.py.py 
    86 .. _lymphography.tab: code/lymphography.tab 
    87  
    88 Here we only give gain ratios to :obj:`bestOnTheFly`, so we don't have to specify a 
    89 special compare operator. After checking all features we get the index of the  
    90 optimal one by calling :obj:`winnerIndex`. 
    91  
    92 ================== 
    93 Add-on Management 
    94 ================== 
    95  
    96 .. index:: add-ons 
    97  
    9816.. automodule:: Orange.misc.addons 
    9917 
    100 ================== 
    101 Server files 
    102 ================== 
    103  
    104 .. index:: server files 
    105  
    10618.. automodule:: Orange.misc.serverfiles 
    10719 
    108 ========= 
    109 `environ` 
    110 ========= 
    111  
    112 .. index:: environment 
    113  
    11420.. automodule:: Orange.misc.environ 
    115  
    116 =============== 
    117 R compatibility 
    118 =============== 
    119  
    120 .. index:: R 
    12121 
    12222.. automodule:: Orange.misc.r 
  • orange/Orange/misc/addons.py

    r7997 r8010  
    11""" 
     2================== 
     3Add-on Management 
     4================== 
     5 
     6.. index:: add-ons 
     7 
    28Orange.misc.addons module provides a framework for Orange add-on management. As 
    39soon as it is imported, the following initialization takes place: the list of 
  • orange/Orange/misc/counters.py

    r7408 r8010  
     1""" 
     2================== 
     3Counters 
     4================== 
     5 
     6.. index:: misc 
     7.. index:: 
     8   single: misc; counters 
     9""" 
     10 
    111class BooleanCounter: 
    212  def __init__(self, bits): 
  • orange/Orange/misc/environ.py

    r7992 r8010  
    1 """\ 
     1""" 
    22================================ 
    33Orange environment configuration 
  • orange/Orange/misc/r.py

    r8008 r8010  
    11""" 
     2=============== 
     3R compatibility 
     4=============== 
     5 
     6.. index:: R 
     7 
    28Conversion of Orange's structure into R objects (with rpy2 package). 
    39 
  • orange/Orange/misc/render.py

    r7701 r8010  
     1""" 
     2================== 
     3Render 
     4================== 
     5 
     6.. index:: misc 
     7.. index:: 
     8   single: misc; render 
     9""" 
     10 
    111from __future__ import with_statement 
    212 
  • orange/Orange/misc/selection.py

    r7560 r8010  
     1""" 
     2================== 
     3Selection 
     4================== 
     5 
     6.. index:: selection 
     7.. index:: 
     8   single: misc; selection 
     9 
     10Many machine learning techniques generate a set different solutions or have to 
     11choose, as for instance in classification tree induction, between different 
     12features. The most trivial solution is to iterate through the candidates, 
     13compare them and remember the optimal one. The problem occurs, however, when 
     14there are multiple candidates that are equally good, and the naive approaches 
     15would select the first or the last one, depending upon the formulation of 
     16the if-statement. 
     17 
     18:class:`Orange.misc.selection` provides a class that makes a random choice 
     19in such cases. Each new candidate is compared with the currently optimal 
     20one; it replaces the optimal if it is better, while if they are equal, 
     21one is chosen by random. The number of competing optimal candidates is stored, 
     22so in this random choice the probability to select the new candidate (over the 
     23current one) is 1/w, where w is the current number of equal candidates, 
     24including the present one. One can easily verify that this gives equal 
     25chances to all candidates, independent of the order in which they are presented. 
     26 
     27Example 
     28-------- 
     29 
     30The following snippet loads the data set lymphography and prints out the 
     31feature with the highest information gain. 
     32 
     33part of `misc-selection-bestonthefly.py`_ (uses `lymphography.tab`_) 
     34 
     35.. literalinclude:: code/misc-selection-bestonthefly.py 
     36  :lines: 7-16 
     37 
     38Our candidates are tuples gain ratios and features, so we set 
     39:obj:`callCompareOn1st` to make the compare function compare the first element 
     40(gain ratios). We could achieve the same by initializing the object like this: 
     41 
     42part of `misc-selection-bestonthefly.py`_ (uses `lymphography.tab`_) 
     43 
     44.. literalinclude:: code/misc-selection-bestonthefly.py 
     45  :lines: 18-18 
     46 
     47 
     48The other way to do it is through indices. 
     49 
     50`misc-selection-bestonthefly.py`_ (uses `lymphography.tab`_) 
     51 
     52.. literalinclude:: code/misc-selection-bestonthefly.py 
     53  :lines: 25- 
     54 
     55.. _misc-selection-bestonthefly.py: code/misc-selection-bestonthefly.py.py 
     56.. _lymphography.tab: code/lymphography.tab 
     57 
     58Here we only give gain ratios to :obj:`bestOnTheFly`, so we don't have to specify a 
     59special compare operator. After checking all features we get the index of the  
     60optimal one by calling :obj:`winnerIndex`. 
     61""" 
     62 
    163import random 
    264 
  • orange/Orange/misc/serverfiles.py

    r7744 r8010  
    11""" 
     2================== 
     3Server files 
     4================== 
     5 
     6.. index:: server files 
     7 
    28Server files allows users to download files from a common 
    39repository residing on the Orange server. It was designed to simplify 
  • orange/doc/Orange/rst/Orange.classification.rst

    r7765 r8010  
    44 
    55.. toctree:: 
    6    :maxdepth: 1 
     6   :maxdepth: 2 
    77 
    88   Orange.classification.bayes 
Note: See TracChangeset for help on using the changeset viewer.