zouter
commented
5 years ago Hi Jonathan!
This file will be mounted inside the docker based on the parameters you give to infer_trajectory. For each method, you can check out these parameters using ?dynmethods::ti_paga and ?dynmethods::ti_paga_tree.
It's true that the difference between these methods is currently not discussed in the documentation. This is on our to-do list!
The difference between ti_paga and ti_paga_tree is that the latter uses the minimal spanning tree, while paga uses a cutoff on the connectivity matrix. The diff between the two files:
105c105
< np.triu(adata.uns["paga"]["connectivities"].todense(), k = 0),
---
> adata.uns["paga"]["connectivities_tree"].todense(),
110c110
< milestone_network = milestone_network.query("length >= " + str(params["connectivity_cutoff"])).reset_index(drop=True)
---
> milestone_network = milestone_network.query("length > 0").reset_index(drop=True)
diegogarcia101
Hello, I am using your excellent package in a project and in your code you reference a parameters file:
"/ti/input/params.json"
Is this file available? I have looked around and cannot find it and it would be most helpful as I am looking at some of your methods in detail. For example, what distinguishes "PAGA" from "PAGA Tree"?
Thanks very much in advance,
Jonathan Greene