-
Notifications
You must be signed in to change notification settings - Fork 55
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
Possible performance increase for Variogram. _calc_groups()
.
#145
Comments
Hi Martin, That looks more than promising. Yeah, I was always aware of that bottleneck, but never found the time to solve it. So I am more than thrilled to implement your solution. I hope to do that soon. If the implementation is urgent for you, you can also file a PR or let me know and I try to hurry up a little bit. Best, |
Hi Mirko, There is no real hurry. I was peeling apart your code as an example for an application meant for internal usage at the TU Wien. I like to have variograms for ASCAT soil moisture data. But there I need some more performance, so I was profiling the code and looking to build in some memoizing functionality for e.g. methods that contain expensive calculations. That is why I came across this and I thought I let you know. Perhaps I can show you this code some time, also to check if credits are due somewhere for if it should go public. I would also be happy to try and turn the above into a PR. Kind regards, |
Yeah, nice. Then I guess the best way to proceed is that you open a PR with your code, which also makes you a collaborator (and co-author of the Zenodo publication). That way credit is attributed clearly and transparently. I can also assist with the PR, if you feel like I can be helpful. |
Yes, and I should list you probably as contributor (or cite your work) on my ASCAT application if it would go public. As I am heavily relying on your knowledge of spatial autocorrelation. |
Just to let you know: There is also a model description paper on GMD, which you can cite if you use the package, DOI: 10.5194/gmd-15-2505-2022 |
I ran into this problem while dealing with a large amount of data. I wonder whether replacing the
for
loop inVariogram. _calc_groups()
. withnumpy.digitize()
might be an option. Not sure if the output is the desired format but it would surely be fast. See below for more details and some benchmark output:Output:
The text was updated successfully, but these errors were encountered: