Erik Garrison

What are the longest sequences you've tested this on? On Sun, May 16, 2021 at 3:44 PM Erik Garrison ***@***.***> wrote: > That's very pragmatic. Other strategies might be more...

I will share some slices of a human chromosome that are from 10kb to 1mbp. On Mon, May 17, 2021, 03:44 Yan Gao ***@***.***> wrote: > The longest one is...

@glennhickey what alignment parameters are you using?

Any workaround (even dropping into non-banded mode when this happens) would be helpful! What do you suggest? Running everything non-banded to avoid this issue would be expensive.

There is now a flag on the result object that indicates if the traceback was OK. It's not propagated to python. On Tue, Oct 13, 2020, 21:15 Roger Volden wrote:...

Unfortunately, I still find cases that cause this error. [fail_smoothxg_block_9338.fa.txt](https://github.com/yangao07/abPOA/files/5380527/fail_smoothxg_block_9338.fa.txt) ``` -> % abpoa -s -r 3 fail_smoothxg_block_9338.fa.txt [simd_abpoa_align_sequence_to_subgraph1] Error in cg_backtrack. (4) ```

Not the most precise definition of what I mean so let me know if it needs clarification.

A pair perfect hash function linking two perfect hashes of the same data with different base hash functions might provide a way to get a set with membership queries in...

Also, definitely test out https://github.com/skarupke/flat_hash_map and https://github.com/greg7mdp/sparsepp. I'd love to hear about any other alternatives. In my application I am mostly limited by lookup speed, and can handle a larger...

The pairPMHF wouldn't be too useful when storing 13-mers, because they are smaller than the mapping that's needed between the PMHFs. But if the keys are very big then the...