We can get a more accurate measure of true test error by evaluating the model fit

We can get a more accurate measure of true test error by evaluating the model fit on this held out set of data. Using the same linear regression model from part 4 (fit to all of the training data), calculate the "true" test error of your predictions based on the newly collected measurements in algaeTest.txt. Is this roughly what you expected based on the CV estimated test error from part 4?