Topics:

• Command line options
• File formats supported
• Output

Primary options:

• -file <name>: location by individuals matrix of SNP values (0,1,2)
• -n <N>: number of individuals
• -l <L>: number of locations
• -k <K>: number of populations
• -rfreq <val>: checks for convergence and logs output every iterations
• -nthreads <val>: use threads

Other options:

• -help: usage
• -label: descriptive tag for the output directory
• -file-suffix: save files with the corresponding iteration as suffix
• -force: overwrite existing output directory
• -seed <val>: value is a real number (read as "double") and sets the seed for GSL
• -compute-beta: computes beta values, given a completely theta fit
• -locations-file: text file of markers for -compute-beta (if not all markers). File is formatted as one SNP index per line, starting from 0.

We support PLINK binary format .bed.

We also support a text format *.012, where SNPs as coded as 0, 1, and 2 and there is no delimiter.

• theta.txt: admixture proportions
• validation.txt: validation likelihoods for the fit. The columns are: iteration number, time elapsed (in seconds), validation likelihood, number of SNPs sampled, and perplexity.