Partition a dataset using an arbitrary custom rule.
The partitioner is configured by passing a custom rule (splitrule) to its constructor. Such a rule is basically a sequence of partition definitions. Every single element in this sequence results in exactly one partition set. Each element is another sequence of attribute values whose corresponding samples shall go into a particular partition.
Notes
Available conditional attributes:
(Conditional attributes enabled by default suffixed with +)
Examples
Generate two sets. In the first set the second partition contains all samples with sample attributes corresponding to either 0, 1 or 2. The first partition of the first set contains all samples which are not part of the second partition.
The second set yields three partitions. The first with all samples corresponding to sample attributes 1 and 2, the second contains only samples with attribute 3 and the last contains the samples with attribute 5 and 6.
>>> ptr = CustomPartitioner([(None, [0, 1, 2]), ([1,2], [3], [5, 6])])
The numeric labels of all partitions correspond to their position in the splitrule of a particular set. Note that the actual labels start with ‘1’ as all unselected elements are labeled ‘0’.
Parameters : | splitrule : list of tuple
enable_ca : None or list of str
disable_ca : None or list of str
count : None or int
selection_strategy : str
attr : str
space : str
postproc : Node instance, optional
descr : str
|
---|