Stuart Lee

@lawremi I'm thinking of refactoring the DeferredGRanges (most likely in the next `plyranges` release) to something where could extend the deferred reading to files other than BAMs - here's what...

@lawremi been going through this again - what do you think should happen in this case: ```{r} fl % select(mapq, qual) ``` If you cache the data using the filter...

Yes that's what I'm saying. I've made it so filter doesn't always cache in the operator-class branch https://github.com/sa-lee/plyranges/commit/3746c5a12d8b7a60fafe556bcc3f04ee73f27792 but was thinking maybe its better to be explicit with filling the...

I guess the issue is when to override and invalidate the cache - I guess if the user supplies arguments that obviously refer to a file operation and are names...

Been thinking about this one lately: is it worth considering moving the IO functions out to their own package and implementing them? We've talked about making all the genomic file...

Glad you managed to get something to work for your needs. When I have more time, I will try to implement a family of `join_nearest_neighbor_*()` functions based on your use...

Thanks, so I definitely think we should include an option for this - but I'm a bit swamped with other things so it might be a while to look at...

I don't think so - I think it just requires updating the reduce_ranges signature to allow for additional arguments to `reduce` via `...`. I should have some time in the...

So `group_by` is really slow when the number of groups gets large and is generally much slower than S4Vectors aggregate compared to group_by + summarise: ```{r} library(plyranges) library(microbenchmark) make_rand_gr %...

see also commit https://github.com/sa-lee/plyranges/commit/9df6fba2989b9893f72bc660b38fdd2c8a0490ca