Reputation: 53
I think this is not a difficult question, but I just got stuck. For two independent datasets (vectors), original values x and predicted values y, I use t-test(x,y,var.equal) to calculate the mean values of these two datasets. How to calculate the difference between the means? |mean of x - mean of y|/|mean of x| or |mean of x - mean of y|/|mean of y|? Thanks for your help.
Upvotes: 0
Views: 105
Reputation: 146
So first asuming this as your data:
x: original values x1,x2,x3...xn
y: predicted values y1,y2,y3...yn
then simply:
VarmeansX <- var(x)
VarmeansY <- var(y)
so you have in VarmeansX and VarmeansY the variance of each data samples, so in order to know what is the difference just do
VarBetween <- VarmeansX - VarmeansY
so you will have a positive or negative number, if the case is the first then the variance in your original values is greater than the predicted values,etc.
Upvotes: 1