[Octave-bug-tracker] [bug #36718] moment, skewness, and kurtosis differences with Matlab

[Rik]

Follow-up Comment #23, bug #36718 (project octave):

I ran 'test kurtosis' and I get a failing test


 ***** assert (kurtosis ([1:5 10; 1:5 10],  0, 2), 5.437731792528892 * [1; 1],
eps)
!!!!! test failed
ASSERT errors for:  assert (kurtosis ([1:5, 10; 1:5, 10], 0,
2),5.437731792528892 * [1; 1],eps)

  Location  |  Observed  |  Expected  |  Reason
    (1)         5.4377       5.4377      Abs err 1.7764e-15 exceeds tol
2.2204e-16
    (2)         5.4377       5.4377      Abs err 1.7764e-15 exceeds tol
2.2204e-16


How did you determine the expected value for this test?  Did you use an
arbitrary precision calculator?  The results for Octave show a difference of
'2' in the last significant digit.


format long 
kurtosis ([1:5 10; 1:5 10],  0, 2)
ans =                                                                         
                                     
   5.437731792528890                                                          
                                                
   5.437731792528890   




    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?36718>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/