GOF problems in version 6.4

announcements (new versions, changes, bugs, installation problems...) related to program PRESENCE

GOF problems in version 6.4

Postby sBorrego » Mon Apr 21, 2014 11:36 pm

Hi,
Trying to run goodness of fit tests in presence v6.4 and presence is returning strange results (e.g. c-hat = 0, extremely high test statistics), not what I was expecting. As a test I tried several models that I already knew the c-hats from previous GOF analyses in earlier versions of presence, and all tests in 6.4 returned strange results (c-hat =0, sky high test stats). I had used the same amount of bootstraps (10k) for all tests.

I then wondered if it was my data/models or was it presence. I ran another GOF test on the same models in an earlier version of presence (v4.3) on another older machine and v4.3 returned "reasonable" results that I would expect (not c-hat=0). In other words, v4.3 is working with the GOF and for some reason v6.4 is not, at least on my two separate computers.

Thoughts? Bug?

Edit: Also the GOF test in v6.4 (using windows 7) required nearly 4 hours to complete 10K bootstraps, whereas the same model and number of bootstraps in v4.3 (using windows vista) only required 1.5 hrs to complete.
sBorrego
 
Posts: 2
Joined: Fri Aug 10, 2012 3:13 pm

Return to software problems/news

Who is online

Users browsing this forum: No registered users and 1 guest