I have a data frame like this (1000000 rows):
A B C
a 0.2 4
b 0.8 7
c 1 8
d 0.2 1
e 0.6 9
I would like to multiply each value in B by a random number taken from C (permutation without replacement) producing a new value x for each row. Then I want to sum all x values obtaining a new row y. Repeat this n times. I don't need the xn columns in my data frame, just a vector with the yn values.
I would get something like this (2 iterations):
A B C x1 x2 .... xn
a 0.2 4 0.2*1=0.2 0.2*4=0.8
b 0.8 7 0.8*8=6.4 0.8*9=0.72
c 1 8 1*7=7 1*1=1
d 0.2 1 0.2*9=1.8 0.2*7=1.4
e 0.6 9 0.6*4=2.4 0.6*8=4.8
y 17.8 8.72 .... yn
Using replicate helps here:
n <- 10
(y <- with(df, replicate(sum(B * sample(C)), n = n)))
# [1] 16.4 16.4 18.0 17.8 14.2 14.2 18.0 20.4 15.2 19.8
If sampling with replacement were an option, it could be made faster by generating a single large matrix of values of C and then using colSums.