search

SCA-IRCM / SingleCellSignalR_v1

R package
26 stars 17 forks source link

clustering - socketConnection #14

Open doctorjb77 opened 4 years ago

doctorjb77 commented 4 years ago

Hello,

Running into issue with clustering the dataset of ~ 5000 cells using 10x scRNA-seq.

clust <- clustering(data = data, n = 10, method = "kmeans") Estimating the number of clusters Error in socketConnection("localhost", port = port, server = TRUE, blocking = TRUE, : all connections are in use

can you suggest any work around for this?

Thanks,

SCA-IRCM commented 4 years ago

Hello, This error comes from the parallel computation of the SIMLR function SIMLR_Estimate_Number_of_Clusters(). You can set your parallel computing parameters (see https://stackoverflow.com/questions/19248598/cant-open-sockets-for-parallel-cluster for example). However if you are looking at the number of clusters in your dataset and you don't want to go down this road, you can set the n.cluster parameter to a random value and see on the t-SNE plot the number of visible clusters, then re-run the clustering() function with this number. You could also use another solution (Seurat or other) for the pre-processing before using the specific functions of SingleCellSignalR.

Thanks for using our package!

SCA