Categ_non_param_cdf<T> is a non-parametric cdf of a categorical (discrete) variable Z. It is defined by n category labels z1,..., zn and the corresponding probabilities. We define the cumulative probability of being in class zj by:
T | the cdf's type value. It can be any ``discrete'' type (e.g. int or bool). It is unsigned int by default. |
z_iterator is a model of Forward Iterator. The z-values z1,..., zn are read from range [z_begin,z_end). The corresponding probabilities are not initialized.
z_iterator and p_iterator are models of Forward Iterator. The z-values z1,..., zn are read from range [z_begin,z_end), and the corresponding probabilities are read starting from p_begin. Since the probabilities are cumulative probabilities, the (cumulative) probability associated to the last class is necessarily equal to 1.
z_iterator is a model of Forward Iterator. The new set of labels can contain more values than the previous one. The probability values corresponding to the former labels are invalidated.