Dear sleepwalk developers, Thanks a lot for providing such nice method. I could install the package but I get the following error when I tried to run:

> sleepwalk([email protected][email protected], [email protected][email protected])
Estimating 'maxdist' for feature matrix 1
Server has been stopped.
Server has been stopped.
Error in app$openPage(useViewer, browser) : 
  Timeout waiting for websocket.
In addition: Warning messages:
1: In file(con, "r") :
  cannot open file 'sleepwalk_canvas.html': No such file or directory
2: In func(req) : File '/favicon.ico' is not found

I know this is probably not a sleepwalk specific error, but I couldn't find a solution for this. Any hints/help on how to fix this issue?

Also, I have a question about the output. Besides using the interactive mode to manually inspect cells that might be "misplaced" on the reduced-dimension space, I would like to systematically find the cells that don't quite fit to the clusters they were originally assigned to. In other words, how would you suggest to use sleepwalk to refine my clustering since I suspect that many of my cells were wrongly assigned to their clusters. I am using Seurat package to reduce dimension and clustering.

Thank you very much, Gustavo

Hello,

After calling the sleepwalk function on a Seurat object, I got this error:

> sleepwalk( as.matrix([email protected][email protected]), as.matrix([email protected][email protected]) )

Estimating 'maxdist' for feature matrix 1
Error in browseURL(str_c("http://localhost:", port, "/", pageobj$startPage),  :
  'browser' must be a non-empty character string

I have loaded the stringr library (containing the function str_c()), and I cannot find the file originating this error. Can I ask if someone had this problem at some point?

Thank you

Running sleepwalk without attaching the package (i.e., NOT specifying library(sleepwalk)) like this works fine:

sleepwalk::sleepwalk(se[email protected][email protected], t([email protected][[email protected],]))

But the moment you select cells with your mouse, it crashed (browser tab closes) and R gives this error:

Error in slw_on_selection(selPoints, 1) : could not find function "slw_on_selection"

Loading the package using library(sleepwalk) solves the issue, but it'd be nice if it weren't necessary.

The example on the web page for comparing two embeddings still uses the old version where both distances are used concurrently. We also need to change the explanation below to say that the same cell always has the same colour in all embeddings

Let say I have e.g. a matrix where I have individuals (cells e.g.) as rows and features as columns, and then run a UMAP on both the ordinary matrix, and the transposed one. Then it would be natural to want to look at the individual UMAP with the default usage (the distances to other individuals), but it would also be interesting to see the features for that individual (and vice versa).

Is it clear what I mean?