Ridge Regression

We are presented with 20 measurements on 3 variables, x1, x2, x3,
and the associated yields Y. The data are as follows:

X1 X2 X3 Y
    1.0000    0.9285   -4.0000   49.9981
    2.0000    2.2792   -4.0000   80.1677
    3.0000    4.3733   -4.0000  118.2350
    4.0000    4.1798   -4.0000  123.2392
    5.0000    4.4580    4.0000   95.5528
    6.0000    7.6342    4.0000  151.0991
    7.0000    7.8252    4.0000  163.4087
    8.0000    8.2308    4.0000  178.8525
    9.0000    9.6716   -2.0000  244.7722
   10.0000    9.4919   -2.0000  254.7109
   11.0000   11.8564   -2.0000  297.7836
   12.0000   12.2685   -2.0000  314.7686
   13.0000   13.6250    2.0000  322.0617
   14.0000   12.9527    2.0000  323.4238
   15.0000   16.5357    2.0000  386.5987
   16.0000   16.4344    2.0000  392.4983
   17.0000   15.0829         0  398.4658
   18.0000   18.4699         0  457.0982
   19.0000   20.2744         0  496.9209
   20.0000   20.6385         0  507.7647

The model being examined is Y = B0 + B1X1 + B2X2 + B3X3 + Error.
It is suggested that ridge regression might be appropriate here.
a) Find the variance inflation factors (VIFs).
b) Produce a ridge trace plot. 
c) Estimate the beta's. Give both the least squares estimates
and the values that you settle on.
d) Predict the yield at X = (x1, x2, x3) = (10, 10, 0).
e) Comment on whether ridge regression was called for in this case.