When run with a small number of reads most assemblers have some combination of settings that is impossible, or which will produce an empty output file or no file at all.
Such problems should be captured and reported, without necessarily ending the whole optimisation process.
If an assembler is crashing every single time, there might be some problem with the input data (like a poorly formatted FASTQ). We therefore want to tally the failures as a proportion of total runs, and if after some number of runs the assembler has never produced a valid output, end the optimisation process and report back to the user that their reads may be corrupted.
When run with a small number of reads most assemblers have some combination of settings that is impossible, or which will produce an empty output file or no file at all.
Such problems should be captured and reported, without necessarily ending the whole optimisation process.
If an assembler is crashing every single time, there might be some problem with the input data (like a poorly formatted FASTQ). We therefore want to tally the failures as a proportion of total runs, and if after some number of runs the assembler has never produced a valid output, end the optimisation process and report back to the user that their reads may be corrupted.