pyrho | Fast inference of fine-scale recombination rates | Machine Learning library

pyrho is compatible with both python 2 and python 3. pyrho makes use of a number of external packages including the excellent [numba](https://github.com/numba/numba), [msprime](https://github.com/tskit-dev/msprime), and [cyvcf2](https://github.com/brentp/cyvcf2) packages. As such it is recommended to install pyrho in a virtual environment. If using [conda](https://conda.io/en/master/) this can be accomplished by running. Once you have set up and activated your virtual environment, you must first install [ldpop](https://github.com/popgenmethods/ldpop). Change to a directory where you want to store ldpop and then run. Note that this will create a directory ldpop. pyrho makes use of [msprime](https://github.com/tskit-dev/msprime), which requires [gsl](https://www.gnu.org/software/gsl/) and [hdf](https://www.hdfgroup.org). pyrho also has a dependency on [openssl](https://www.openssl.org). If you do not have these installed, these can be installed using. sudo apt-get install libssl-dev. git clone https://github.com/popgenmethods/pyrho.git pyrho pip install pyrho/. pyrho make_table --samplesize <n> --approx --moran_pop_size <N> \ --numthreads <par> --mu <mu> --outfile <outfile> \ --popsizes <size1>,<size2>,<size3> --epochtimes <breakpoint1>,<breakpoint2>. pyrho hyperparam --samplesize <n> --tablefile <make_table_output> \ --mu <mu> --ploidy <ploidy> \ --popsizes <size1>,<size2>,<size3> --epochtimes <breakpoint1>,<breakpoint2> \ --outfile <output_file>. pyrho optimize --vcffile <data> --windowsize <w> --blockpenalty <bpen> \ --tablefile <make_table_output> --ploidy <ploidy> --outfile <output_file> \ --numthreads <par>. pyrho compute_r2 --tablefile <make_table_output> --samplesize <n> \ --quantiles 0.25,0.5,0.75 --compute_mean.