-
Notifications
You must be signed in to change notification settings - Fork 8
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
Feature: Impulsive metrics (kurtosis and energy windowed SPL RMS) #143
Conversation
…s, updated gitignore
* updated tests to include relative paths to test data, allowing them to run anywhere * corrected issue related to non-zero bin overlaps
* updated tests to include relative paths to test data, allowing them to run anywhere * draft metrics commit * add impulsive metrics and associated tests * small corrections/refactor * added kurtosis units, added test for using apply multiple for kurtosis, updated gitignore * updated docstrings * clean up test_impulsive_metrics.py
fix invalid escape
add numpy import
added numpy import
removed unnecessary import
Small update here: after adding the optional dependencies, all tests pass. |
Apologies for the messy commit history, and I should note that this branch also contains the bin overlap fix related to #127. Again, if you want me to more clearly separate out all of these changes, please let me know and I can submit pull requests from simpler/cleaner branches. |
Hi, no worries! Thank you so much for all the nice work, it looks great! I am now busy with some other work but I will definitely check it out in in two weeks :) Sorry for the delay. |
Absolutely no rush on my end - happy to work through this whenever it makes sense. Thanks! |
Thank you for merging this pull request, I'm seeing that one of my tests are failing in CI. I believe this is because I'm handling paths in a windows specific way, I'm working on a pull request that should resolve this. |
Hello,
Finally ready to submit this, the changes can be summarized as:
In terms of new math, the primary things are the kurtosis and energy window calculations. For kurtosis, I explored the best implementation and documented it here. For the energy window calculation, I followed the standard method from the most often cited reference. In both cases I added tests which compare against known/expected results for each. Additionally, I built a comprehensive test of all of the new metrics using some example pile driving acoustic data and previously analyzed results.
In terms of testing, everything is passing with the exception of
TestPlots.test_summary_plot
andTestDataset.test_generate_dataset
.Here are the two exceptions from the failed tests:
Please let me know if you would like me to split any of this out and submit separate pull requests, as I am realizing this might be a lot for a single pull request. For example, if you are only interested in including kurtosis, I would definitely understand, as many of my edits specifically relate to performing analysis of anthro impulsive sources.
Again, it was been really fun to work with and learn this package, thanks!