Double scalars overflow within for loop

Question

I know that overflow double scalars occurs when we exceed the range allowed for a dtype. In my case I\'m trying this

l=l+pow((y[i,]-(np.dot(w,x[i,]))),2)
<