-
Notifications
You must be signed in to change notification settings - Fork 5
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
A few questions on rawhash #1
Comments
Hi @hasindu2008, Thanks for your interest in using RawHash! Interesting that you got this error. We were not getting this linking error. I just changed the Makefile to link pthread and standard C++ libraries. Could you pull the newest changes and try again? For your questions:
We are planning to provide an easy-to-usage support for the recent chemistries soon though. This includes the easy parsing of the recent k-mer model released my ONT (with fewer columns) and optimizing the parameters for the recent chemistries. Quantization and other parameters (-k, -e, -q, -l) may need to be slightly changed to optimize accuracy and performance.
|
Did you forget to commit? I cannot see a recent change in the makefile. |
You are absolutely right! Probably due to the networking issue, the push did not go through on our end. The following commit is now pushed: 9099df0 |
Unfortunately I am getting the same error.
Do you use some compiled binaries for some libraries? This at a glance looks like an incompatibility of some precompiled binaries with the particular compiler version I am using. |
Hi @hasindu2008, You are correct. Due to incorrect parsing of your system information, RawHash was probably downloading the incompatible POD5 libraries. I have been working on making the Makefile more flexible to mitigate these issues. I have made some changes to Makefile. You can now either 1) disable compiling the POD5 libraries if you do not need them or 2) provide the path to your own POD5 include and lib directories. Please check https://github.com/CMU-SAFARI/RawHash#compiling-with-hdf5-slow5-and-pod5 for more detailed information. For sanity check, could you try to compile by disabling POD5 as follows: make NOPOD5=1 Additionally, I have changed the implementation for parsing the k-mer model files. RawHash should now correctly parse both R9 and R10 k-mer model files. For R10, you may need to specify Note: We have not tested RawHash for R10 to provide the best possible parameter settings. The default preset parameters for R10 will be available soon. |
Hi @hasindu2008, Are you still experiencing the same problem? (if this is the case, please for now compile it with the On a separate note, we have also added the SLOW5 support to RawHash although I could not extensively test it in a wide range of scenarios. I would appreciate your feedback if you notice any issues there as well. |
Hi |
Fantastic! Thanks for adding the S/BLOW5 support. I was able to quickly compile using Have you done a large benchmark - interested to see how much Would you mind if I added rawHash to the list here? Suggestions: |
Excellent! Glad to hear that things are working fine now. I haven't done a large benchmark yet to identify the potential bottlenecks. I will keep in mind your suggestion as currently there is definitely some room for improving the thread utilization of RawHash. To implement your suggestion, we could assign more threads to the IO part rather than allocating almost all of them for mapping. Thanks a lot for the pointers and the tip regarding Of course, we would be happy to see RawHash to be listed there. |
Done. Let me know if you had any difficulties in using slow5lib and if you have any suggestions. Nowadays, it is hard to find anyone writing things in C. Feel free to close this issue. |
Hi,
rawhash looks fantastic and I wanted to give it a try. But I got an error in the compilation process.
I have attached the log from the
make
command for your reference. Could you have a look and let me know some tips?log.txt
Also, I have a few questions.
The text was updated successfully, but these errors were encountered: