-
Notifications
You must be signed in to change notification settings - Fork 22
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
Problem in buildNhoodGraph on a large dataset #290
Comments
Hi @moinfar - two things to start with: 1) you don't need to run the nhood refinement with 100% of the data - that is entirely redundant and a sure-fire way to kill all of your statistical power, 2) use the graph-based nhood refinement and spatial FDR correction - it's orders of magnitude faster. Finally, there is a milopy implementation maintained by @emdann that you may find useful: https://github.com/emdann/milopy |
Hi @MikeDMorgan, Thanks for your reply.
The difference between the two runs is that I subset the whole data (initial adata) to 10% in the working one, while I use the whole data in the other. The latter results in the given error. Am I missing something?
Thanks for mentioning the milopy implementation. I will take a look at it. |
Hi @moinfar I see - sorry from your post I inferred you set p=1.0 - the original nhood refinement time complexity for the jth nhood is I haven't seen that particular error before - it suggests there are NA of Inf values in your |
Did you solve the error @moinfar ? |
@ManarHashemTaha |
Oh no, I have the same issue here. My data includes 221k cells |
Describe the bug
Hi,
First, thanks for providing and maintenance of the package.
I am using Milo on a dataset of 588k cells. I am trying to run Milo in Jupyter environment using the rpy2 interface (like here).
Everything is fine when I subset the cells to 10% (59k cells). However, when I run it on the whole data with 588k cells, I face an error on the buildNhoodGraph function. It looks like an overflow problem, but I am not sure. Unfortunately, the data is too big for me to upload. But please let me know if I can provide anything else.
Minimum code example
Minimum example to reproduce the error
Full error traceback
Session info
Output of
sessionInfo()
P.S
Some minor differences between 10% and 100%:
da_results <- testNhoods(...)
:The text was updated successfully, but these errors were encountered: