Generate diagonal matrix from regression coefficient

Question

I am trying to generate a diagonal matrix using a linear regression coefficient. First I generated an empty matrix. Then I extract the coefficient from the regression model. Here's my code:

P = np.zeros((ncol, ncol), dtype = int)
intercep = np.zeros((1, ncol), dtype = int)

my_pls = PLSRegression(n_components = ncomp, scale=False)
model = my_pls.fit(x, y)

#extract pls coeffeicient:
coef = model.coef_
intercep = model.y_mean_ - (model.x_mean_.dot(coef))

P[(i-k):(i+k), i-k] = np.diag(coef[0:ncol])

But I got zero matrices after running the code. Can anyone please help me out with how to get the diagonal matrix from the regression coefficient?

Answer 1

Not sure why you need to declare P .

You can get diagonal matrix with zeros directly from the 1D list/vector using numpy.diag

x=[3,5,6,7]
numpy.diag(x)

Output:

array([[3, 0, 0, 0],
       [0, 5, 0, 0],
       [0, 0, 6, 0],
       [0, 0, 0, 7]])

For your case, try P=np.diag(coef)