Joyjit Daw
Joyjit Daw
The cudaaligner API only supports `ATCG` alphabet, but the API doesn't check for this in input strings right now, leading to undefined behavior.
1. each aligner batch to take in max memory and max ref/query sizes and determine how many how alignments can be performed in the batch. 2. provide api to check...
Suggestions from @ahehn-nv - 1. ``` You could avoid the construction of the std::string by directly operating on the stream: std::ostream& operator
The `version.cpp.in` file should be configured once as part of the base module, and all the modules should simply call into the `version.hpp` header. currently that file is being separately...
Right now the macros for banded alignment like CUDAPOA_BANDED_MATRIX_RIGHT_PADDING are not shared between the kernel code and other sources that determine sizes for bands (such as in batch.cu). For now...
The reference to the stream object is getting lost somehow in CudaAlignerBatch object, because of which stream and batch are sometimes getting destroyed out of order. Explicit ref increment/decrement resolves...
Our CI scripts currently written in `bash`, and use a lot of env vars and positional arguments. But this setup is ripe for bugs and errors. It's important to upgrade...
Python doc generation only happens successfully when the bindings have been built. However the current sphinx command doesn't fail when bindings aren't built. This issue is to investigate the right...
Currently we only have unit tests for cudamapper, but nothing that runs the bianry on a dataset end to end. This should be added to gpuCI runs as well, for...
Wrap the C++ error enums into an error class that has a to `__str__` function and can be thrown from within the binding classes