在R中的一列上分布多个列

Question

Sorry because this question has been asked several times, but I'm still having trouble wrapping my head around this problem.

So I have a dataframe, of the form:

ID Val Type
1  10    A
2  11    A
2  10    C
3  10    B
3  12    C
4   9    B

It's not much help but you can use

library(tidyr)
test <- data.frame(ID = c(1,2,2,3,3,4), 
                   Val = c(10,11,10,10,12,9), 
                   Type = c('A', 'A', 'C', 'B', 'C', 'B'))

I would like to split it to obtain:

ID A.Type B.Type C.Type A.Val B.Val C.Val
1    1     0       0     10    0     0
2    1     0       1     11    0    10
3    0     1       1      0   10    12
4    0     0       0      0    9     0

I know how to get columns 1:4 using:

table(test[, c(1, 3)]) %>% as.data.frame() %>% spread(Type, Freq)

It's the last three I need help with because in the actual data-frame values are continuous and table can not be used.

Answer 1

You are trying to reshape your data with multiple value variables where the ones are actually implicit, so in order to get the type_... columns, you will need to create a new type variable with ones and then use dcast from data.table package:

library(data.table)
setDT(test)
dcast(test[, type := 1][], ID ~ Type, value.var = c("type", "Val"),fill = 0)
#    ID type_A type_B type_C Val_A Val_B Val_C
# 1:  1      1      0      0    10     0     0
# 2:  2      1      0      1    11     0    10
# 3:  3      0      1      1     0    10    12
# 4:  4      0      1      0     0     9     0

Or you can use reshape from base R, where NA has to be manually replaced:

test$type = 1
reshape(test, idvar = "ID", timevar = "Type", direction = "wide")

#   ID Val.A type.A Val.C type.C Val.B type.B
# 1  1    10      1    NA     NA    NA     NA
# 2  2    11      1    10      1    NA     NA
# 4  3    NA     NA    12      1    10      1
# 6  4    NA     NA    NA     NA     9      1