Hi,
On 04/17/2015 09:42 AM, Hervé Pagès wrote:
Hi,
I think we should just expand the Rle internally. That will produce
a numeric vector of the length of the GRanges i.e. it will be the
same size as the start and end components of the GRanges object itself.
No big deal at all.
I'll make that ch
Sorry, I was confused. You're right, there's not much point in keeping it
compressed.
On Fri, Apr 17, 2015 at 10:14 AM, Hervé Pagès wrote:
> On 04/17/2015 10:00 AM, Michael Lawrence wrote:
>
>> Is that the case here? He has an Rle as an mcol in the GRanges, so in
>> general expanding it will not
On 04/17/2015 10:00 AM, Michael Lawrence wrote:
Is that the case here? He has an Rle as an mcol in the GRanges, so in
general expanding it will not align with the other components.
Not sure what you mean. Can you give an example?
H.
On Fri, Apr 17, 2015 at 9:42 AM, Hervé Pagès mailto:hpa...
Is that the case here? He has an Rle as an mcol in the GRanges, so in
general expanding it will not align with the other components.
On Fri, Apr 17, 2015 at 9:42 AM, Hervé Pagès wrote:
> Hi,
>
> I think we should just expand the Rle internally. That will produce
> a numeric vector of the length
Hi,
I think we should just expand the Rle internally. That will produce
a numeric vector of the length of the GRanges i.e. it will be the
same size as the start and end components of the GRanges object itself.
No big deal at all.
I'll make that change.
H.
On 04/17/2015 09:00 AM, Michael Lawren
Ideally it should be supported, but it would take some work as the coverage
stuff is all in C. Could you give more details on your use case? For
example, if you already have a range for every position on the chromosome,
you could just extract the score column. I'm guessing it's more complicated
tha
Dear all, I'm puzzled by the following behaviour:
Given
n <- 10
gr <- GRanges(seqnames=Rle('A', n),
ranges=IRanges(1:n, width=1),
score=Rle(5,n))
If I do
coverage(gr,weight='score')
I get
Error in .normarg_shift_or_weight(weight, "weight",