R model variable score

I built a linear regression model in R and tried to get the contribution amount (not the coefficient) of each of the explanatory variables (independent variables) (i.e. x1, x2, x3). Question:

How can I get such amount for the fitted data. That is numeric value contributed by each x (x1, x2, x3) for each observation.
How can I get the same amount for new (unseen) data

df = data.frame(y, x1, x2, x3)
mod = lm(y ~ x1 + x2 + x3, data = df)

Solution

As @Ali suggested, I think that summary(mod) does answer your question. Let me give a bit more explanation. Since you do not provide your data, I will use the built-in iris data as an example.

mod = lm(Sepal.Length ~ ., data=iris[,1:4])
summary(mod)$coefficients
               Estimate Std. Error   t value     Pr(>|t|)
(Intercept)   1.8559975 0.25077711  7.400984 9.853855e-12
Sepal.Width   0.6508372 0.06664739  9.765380 1.199846e-17
Petal.Length  0.7091320 0.05671929 12.502483 7.656980e-25
Petal.Width  -0.5564827 0.12754795 -4.362929 2.412876e-05

Notice the column labeled "Estimate". Those are the model coefficients. Just to be really explicit, let's go through an an example of how they relate to the prediction.

iris[1,]
  Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1          5.1         3.5          1.4         0.2  setosa

predict(mod, newdata=iris[1,])
       1 
5.015416

OK, so if we predict the first row using the model, we get the answer 5.015416. How did that come from the coefficients?

 1.8559975 +            ## Intercept
 0.6508372 * 3.5 +      ## Sepal.Width
 0.7091320 * 1.4 +      ## Petal.Length
-0.5564827 * 0.2        ## Petal.Width
[1] 5.015416