Similarly to read.preclustered.datasets, the unique commands needs to be removed from all the following code (taken from read.data function). We have already don't that for read.preclustered.datasets previously. The structure of initialize.project is that each dataset has a value of origins and experiments which can be duplicated. Applying unique onto these vectors is problematic.
dataset.labels <- rep(paste(unique(environment$origins)[environment$datasets ==
dataset], " (", unique(environment$experiments)[environment$datasets == dataset],
")", sep = ""), ncol(measurements))
origins <- rep(unique(environment$origins)[environment$datasets == dataset],
ncol(measurements))
experiments <- rep(unique(environment$experiments)[environment$datasets == dataset],
ncol(measurements))
See example of how I run initialize.project:
combined <- initialize.project(datasets = c("TbetWT_Exp.1","TbetWT_Exp.2","TbetKO_Exp.1","TbetKO_Exp.2"),
origins = c("Tbet NKp46 WT cells","Tbet NKp46 WT cells","Tbet NKp46 KO cells","Tbet NKp46 KO cells"),
experiments = c("Replicate 1","Replicate 2","Replicate 1","Replicate 2"),
data.path = '/data/magenay/ILC/data',
work.path = '/data/magenay/ILC',
marker.genes = markers)#,clear.history=T
Similarly to read.preclustered.datasets, the
uniquecommands needs to be removed from all the following code (taken from read.data function). We have already don't that for read.preclustered.datasets previously. The structure ofinitialize.projectis that each dataset has a value oforiginsandexperimentswhich can be duplicated. Applying unique onto these vectors is problematic.See example of how I run initialize.project: