Update read.bismark documentation about bsseq HOT 3 CLOSED

Was this ever fixed? Bismark now (?) has several types of output files. We should update docs to include a discussion of this and also add unit testing for parsing by just reading in say 10 lines or so.

from bsseq.

Yes, I think these were covered in #4. So I think this can be closed.

Incidentally, that link includes my benchmarking of data.table::fread() and readr::read_tsv() on gzipped data that we discussed yesterday. I think the data still stand, but I'm more open to the interpretation that readr::read_tsv() is better for bsseq because it natively handles gzipped files (they are uncompressed in memory rather than to disk; http://blog.rstudio.org/2015/04/09/readr-0-1-0/).

from bsseq.

One issue with our compute cluster is that disk access is slow. If
benchmarking was done on laptop, highly likely that we may see different
results on JHPCE, because readr avoids writing a file.

On Fri, Mar 11, 2016 at 3:21 PM, Peter Hickey [email protected]
wrote:

Yes, I think these were covered in #4
#4. So I think this can
be closed.

Incidentally, that link includes my benchmarking of data.table::fread()
and readr::read_tsv() on gzipped data that we discussed yesterday. I
think the data still stand, but I'm more open to the interpretation that
readr::read_tsv() is better for bsseq because it natively handles gzipped
files (they are uncompressed in memory rather than to disk;
http://blog.rstudio.org/2015/04/09/readr-0-1-0/).

—
Reply to this email directly or view it on GitHub
#1 (comment)
.

from bsseq.