I am trying to use proDA for a over-drop out dataset of mass-spectrometry analysis, and once i use the data for model fitting, it turns out a error
> fit = proDA(as.matrix(dfFilter), design = condiction)
Error in if (sign(objective_fun(min_var)) == sign(objective_fun(max_var))) { :
missing value where TRUE/FALSE needed
I have test that the data input, with treatments as columns and proteins as rows. In addition, I have filtered all the rows with no real observations (before replace zeros with NAs) :
> any(apply(dfFilter, 1, sum) <= 1e-8)
[1] FALSE
I wonder is there any step I did wrong? or just the data are too many zeors to regression?
I am trying to use proDA for a over-drop out dataset of mass-spectrometry analysis, and once i use the data for model fitting, it turns out a error
I have test that the data input, with treatments as columns and proteins as rows. In addition, I have filtered all the rows with no real observations (before replace zeros with NAs) :
I wonder is there any step I did wrong? or just the data are too many zeors to regression?