Your task is to write a function that takes in a 1-d Numpy array of actual initial values, a numeric weight scaler (between 0.9 and 0.99) and the number of previous time steps for prediction k, and return a numpy array of forecasted values for all values in that array starting with the k+1st one (we need to use the first k actual values to forecast the first observation). The idea here is that the most recent observation takes on a higher weight than less-recent observations in the data. Please use Numpy vectorization methods to solve the problem. .