The rdpout file is designed to be compatible with output from the RDP Naive Bayesian Classifier command-line version. It is tabbed text with a varying number of fields. One line is given for each query sequence.
The first field in a line is the query sequence label.
The second field is always empty, i.e. the query label is always followed by two tabs.
The taxonomy prediction follows the empty field. Each rank has three fields: the taxon name, the rank name and the confidence value, for example:
Halobacteria class 0.78
The first rank is always:
Root rootrank 1.0