-
Notifications
You must be signed in to change notification settings - Fork 77
Add option to compute_mutation_times
to sample times uniformly along the edge.
#849
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
Comments
Just caught up on the RNG discussion here on GPL vs. MIT. As long as the uniform sampling is implemented in In other words, I'm a bit confused what the |
Also, what is the current behavior for a mutation above a node with no parents? And what do we propose we change it to under pseudorandom uniform sampling? |
Here's an example of computing these things using the Python API and making a histogram of all mutation times. It's relatively efficient.
|
I think the only thing we can do is to put the times equal to the time of the node below them. |
I agree that we'll probably need randomness in tskit at some point and using the RNG that you mention above is probably the way to go. I'm with @brianzhang01 and the moment though, in wondering whether this is something we want to take on right now. I'm happy either way though. @petrelharp, do you see a use for this random version of compute_mutation_times? |
No, I'm with @brianzhang01 - if someone needs to do this, the python version is easy and efficient. |
Cool, we'll shelve this for now then. |
Uh oh!
There was an error while loading. Please reload this page.
When
compute_mutation_times
was added we discussed randomly assigning times rather than the current method of using the midpoint (#513 (comment)). We decided to leave the source of randomness as an input to avoid setting up an MIT-license compatible RNG in tskit. However, since then we also have discussed the need of randomness for splitting polytomies (#815 (comment)) and it seems that https://www.pcg-random.org/download.html may fit the bill. I suggest we attempt to add an RNG to tskit for mutation times and see how it goes. Any thoughts? I'd be happy to take this on if there is consensus.This came up in tskit-dev/msprime#977
The text was updated successfully, but these errors were encountered: