Currently, the parallelization is done by distributing the input files over the MPI tasks. However, when the number of input files is less than the number of tasks, the code aborts. For instance, the full meraxes dataset is in one file and therefore can not be converted in parallel currently.
meraxes conversion also takes the longest per GiB and the parallel conversion would be a good thing
Currently, the parallelization is done by distributing the input files over the MPI tasks. However, when the number of input files is less than the number of tasks, the code aborts. For instance, the full
meraxesdataset is in one file and therefore can not be converted in parallel currently.meraxesconversion also takes the longest perGiBand the parallel conversion would be a good thing