The prerequisite for plotting multiple datasets from a output file created with selected_output is that the different datasets are separated according the format of gnuplot, i.e. datasets are separated by two blank lines. Other separation of datasets within a file cannot be handled via the selected_graph statements and should be treated via post-processing.


To add the gnuplot format between datasets, the following statement can be used (see punch):


punch <:int:Selected Output Number> gnuplot [<:str:dataset label>] 



The number of datasets and the default keys for the different curves are defined based on the options given in the Plot Line Options.



Definition of labels of different data sets


The number of datasets and the keys are defined by the number of indexes that are defined for the selected_output. Indexes are assigned to the selected_output by the statements (see punch):


punch <:int:Selected Output Number> gnuplot [<:str:dataset label>] 


or


punch <:int:Selected Output Number> index [<:str:dataset label>] 


If no dataset label is defined, the default label "Dataset i" (with i the ith index) is assigned.