some bugs in testing "1. A basic walkthrough of 3k PBMC cells starting from a pre-computed Seurat object." about swne

[Shiywa]

emmmmm , when I tried the other tutorial ,I found a parameter in EmbedSWNE called dist.use. When I set it in cosine, the tutorial is OK. When I set it in IC, it called error.
In fact, I don't really understand the principle in it.

[yanwu2014]

Thanks for the heads up, it should be fixed!

So the dist.use is setting the metric used to embed the factors. Different metrics, like cosine distance or information coefficient (IC), will have different advantages and disadvantages. We find that setting dist.use = cosine works very well in most cases.

Let me know if that helps!

[yanwu2014]

Closing this thread if there are no more comments!