Suppose I have a model like the following:
$\hat{y} = 100 + 250 \log{x_{1}} + 75 x_{2} + 80 x_{3} + 105 \log{x_{4}}$
If an observation in my validation set had $x_{1} = 50, x_{2} = 40, x_{3} = 30,$ and $x_{4} = 20$, how could I put those into my model to get a predicted value. The response, $y$, is in normal units, while two of the regressors, $x_{1}$ and $x_{4}$, had the logarithm applied to them.