Closed denvercal1234GitHub closed 1 year ago
prepData
, argument panel_cols
defaults to having a marker_class
column in the panel
that contains marker classes (type, state, none). In your example, this doesn't exist. Consequently, all markers are assigned a class of none.cluster
in turn defaults to using marker of class type, but there aren't any, hence the error. So there are 3 solutions: either 1) specify marker classes in the input panel
in prepData
, and set the corresponding column in panel_cols
(preferred), or 2) set marker classes after the SCE has been constructed via rowData(sce)$marker_class <- ...
or 3) pass the markers to use directly to cluster
's features
argument. 1) Is best here because many of CATALYST
's functions default to using type
or state
markers (it depends), and if these aren't set you'll have to always pass them manually to a given function...
Hi there,
Thanks for the tool.
I did the following but when I checked
rowData(F37_channelCD3CD8_fcs_data_sce)$marker_class
, all were "none" with Levels being still "type", "state", and "none."And,
cluster
returnedError in .get_features(x, features) : No features matched the specified marker class.
Would you mind letting me know how to proceed?
Related = #84 #335