[Scipy-tickets] [SciPy] #1799: Bug in linregress() for masked arrays

SciPy Trac scipy-tickets@scipy....
Sat Dec 22 11:20:44 CST 2012


#1799: Bug in linregress() for masked arrays
-------------------------+--------------------------------------------------
 Reporter:  bulli92      |       Owner:  rgommers   
     Type:  defect       |      Status:  new        
 Priority:  normal       |   Milestone:  Unscheduled
Component:  scipy.stats  |     Version:  0.11.0     
 Keywords:               |  
-------------------------+--------------------------------------------------
 The stats.linregress() and stats.mstats.linregress() functions give
 inconsistent results.

 mstats.linregress() returns different values for the p-value and the
 standard error, which seems to be due to a different calculation of the
 t-statistics.

 I attach a unittest routine snipped that can be used to reproduce the
 error.

 Is mstats module currently actively further developed? I checked also
 consistencies between stats and mstats for other routines and could
 implement some further unittests if needed. The current unittests do not
 really cover masked array testcases as all arrays used are standard numpy
 arrays in tests/test_mstats_basic.py

-- 
Ticket URL: <http://projects.scipy.org/scipy/ticket/1799>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.


More information about the Scipy-tickets mailing list