The GenePattern site has example data for use in the original HAPSEG-ABSOLUTE workflow. Perhaps you can start with this.

• For an overview of the original published workflow, using Affy SNP data, see the Using HAPSEG and ABSOLUTE in GenePattern page. The CGA group provides an example dataset for download to use in the workflow. Although data for twenty samples are provided, two are sufficient for the workflow.

ABSOLUTE.summarize and ABSOLUTE.review are an integral part of the ABSOLUTE workflow. It should be clarified that ABSOLUTE provides a number of mathematically computed models of purity and ploidy that it ranks for you based on the combined likelihoods stemming from copy ratios data, mutation data (if provided), and matched reference karyotype. Researchers are expected to manually review the provided solutions and select the most reasonable one, which isn't necessarily the top ranked solution provided by ABSOLUTE. For example, if a sample is from a solid tumor versus metastasized tumor, you probably have an idea of the expected tumor fraction of the sample, e.g. higher for the solid tumor sample than the metastasized sample, that then you would keep in mind when weighing the provided solutions. You modify ABSOLUTE.summarize's calls file to override the top ranked solution with your choice of solution and provide this to ABSOLUTE.review for finalized results.