Closed magicDGS closed 6 years ago
I think it would be a good idea to fail with a reasonably descriptive error if the _SUCCESS file is not present by default. And then maybe have a --force option that changes the error into a warning?
Cheers Rupert
2018-06-06 9:22 GMT+02:00 Daniel Gómez-Sánchez notifications@github.com:
Assigned #88 https://github.com/robmaz/distmap/issues/88 to @robmaz https://github.com/robmaz.
— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/robmaz/distmap/issues/88#event-1665402302, or mute the thread https://github.com/notifications/unsubscribe-auth/Ad_FfPDhAjH6kKZNj4TgnwPo_T-8pSSkks5t54MfgaJpZM4UcHfC .
Version 1.4.1 of ReadTools have this patch.
@robmaz - I found in our wiki help for mapping on the cluster that your script checks for the
_SUCCESS
file before download from the cluster withReadTools
. It is quite easy to implement this check there and avoid thehadoop fs -ls /${pref}_myname/fastq_paired_end_mapping_bwa/_SUCCESS
command for the user.In addition, because this could have some problems to download files if the
_SUCCESS
is not present, we can have an advance argument to disable this check (e.g., download the data from something that fails except in one block, identifying how many reads were lost and evaluating if it is worthy to remap).Tell me what do you think about this - it is really easy to implement as a bug fix, because it is true that downloading a not successful job is not desirable.