Adding grouped mean values to column in data frame [duplicate]

Answer 1

Using data.table

library(data.table)
setDT(mytestdata)[, c("MeanValue", "Residual") := {m= mean(Value);list(m, Value-m)}, by=list(Insertion, Unit, Channel)]

mytestdata

#        Insertion Measurement Unit Channel Value MeanValue     Residual
#       1:         1           1   A5      10  9.41  9.403333  0.006666667
#       2:         1           1   A5      11  9.51  9.470000  0.040000000
#       3:         1           1   A5      12 10.59 10.850000 -0.260000000
#       4:         1           1   A5      13  9.45  9.476667 -0.026666667
#       5:         1           2   A5      10  9.42  9.403333  0.016666667
#       6:         1           2   A5      11  9.03  9.470000 -0.440000000
#       7:         1           2   A5      12 10.62 10.850000 -0.230000000
#       8:         1           2   A5      13  9.39  9.476667 -0.086666667
#       9:         1           3   A5      10  9.38  9.403333 -0.023333333
#      10:         1           3   A5      11  9.87  9.470000  0.400000000
#      11:         1           3   A5      12 11.34 10.850000  0.490000000
#      12:         1           3   A5      13  9.59  9.476667  0.113333333
#      13:         2           1   A5      10 12.10 12.100000  0.000000000
#      14:         2           1   A5      11 11.28 11.280000  0.000000000
#      15:         2           1   A5      12 12.95 12.950000  0.000000000

Answer 2

You can use ave to calculate the groupmeans:

df$MeanValue <- with(df, ave(Value, Insertion, Unit, Channel, FUN = mean))

Then calculate the residuals:

df$Residual <- df$Value - df$MeanValue
df
#   Insertion Measurement Unit Channel Value MeanValue     Residual
#1          1           1   A5      10  9.41  9.403333  0.006666667
#2          1           1   A5      11  9.51  9.470000  0.040000000
#3          1           1   A5      12 10.59 10.850000 -0.260000000
#4          1           1   A5      13  9.45  9.476667 -0.026666667
#5          1           2   A5      10  9.42  9.403333  0.016666667
#6          1           2   A5      11  9.03  9.470000 -0.440000000
#7          1           2   A5      12 10.62 10.850000 -0.230000000
#8          1           2   A5      13  9.39  9.476667 -0.086666667
#9          1           3   A5      10  9.38  9.403333 -0.023333333
#10         1           3   A5      11  9.87  9.470000  0.400000000
#11         1           3   A5      12 11.34 10.850000  0.490000000
#12         1           3   A5      13  9.59  9.476667  0.113333333
#13         2           1   A5      10 12.10 12.100000  0.000000000
#14         2           1   A5      11 11.28 11.280000  0.000000000
#15         2           1   A5      12 12.95 12.950000  0.000000000

Or you could use dplyr

library(dplyr)

df %>% group_by(Insertion, Unit, Channel) %>% mutate(MeanValue = mean(Value), Residual = Value - MeanValue)