Question

How to rewrite the following functions using only numpy array operations (universal functions, aggregations, boolean indexing) without using loops

def mean_squared_error (v , p ):

n = len ( v [0])

result = 0

for i in range ( n ):

result += ( v [0][ i ] - p [0])**2 + ( v [1][ i ] - p [1])**2

result = result / n

return result

I tried doing it like this but nothing printed out

def mean_squared_error2 (v, p ):

v=np.array(v)
n = v.shape[1]
result = np.sum((v[0,:]- p[0])**2 + (v[1,:] - p[1])**2)
print(result)
return result/n

Answer 1