From brodgar@lists.highstat.com Wed Feb 28 20:29:32 2007 From: brodgar@lists.highstat.com (brodgar@lists.highstat.com) Date: Wed, 28 Feb 2007 16:59:32 -0330 Subject: [brodgar]DFA example Message-ID: Hi Alain, Hi everybody,=20 I am a brand-new user of Brodgar, and my apologies in advance if my = question is trivial.=20 I am still exploring the software and digging into the manual and online = examples, but found something that looks odd.=20 I was trying to replicate the results for the DFA Nephrops example in = the website (e.g. Table 1), and I have been unable of doing it. I = normalized the variables and run the analyses but the AIC values I get, = though in the same order of magnitude are still clearly different. For = example, the model with two common trends gives me an AIC around 715 = (like the one with only one common trend in the website example), and = adding more common trends does not bring it below 700. furthermore, the = model with 11 individual trends gives me an AIC around 670, much lower = than the 974 in the example (i.e. my conclusion from this should be 11 = individual trends are better than 2 common trends instead as the other = way around as stated in the example). My basic question is, what am I = doing wrong? Were the data transformed in some way before running the = analyses? What am I missing here? My apologies in advance if this question is too silly, but I want to = start with right foot.=20 Thank you very much,=20 Mariano ************************************** Mariano Koen-Alonso Northwest Atlantic Fisheries Centre Fisheries and Oceans Canada 80 East White Hills Road PO Box 5667 St. John's, Newfoundland A1C 5X1 CANADA Centre des p=EAches de l'Atlantique Nord-Ouest P=EAches et Oc=E9ans Canada 80 East White Hills Road CP 5667 St. John's, Newfoundland A1C 5X1 CANADA Phone: +1(709)772-2047 Fax: +1(709)772-5315 E-mail: Koen-AlonsoM@dfo-mpo.gc.ca **************************************=20 From brodgar@lists.highstat.com Wed Feb 28 23:40:56 2007 From: brodgar@lists.highstat.com (brodgar@lists.highstat.com) Date: Wed, 28 Feb 2007 23:40:56 +0000 Subject: [brodgar]Re: DFA example In-Reply-To: References: Message-ID: <7.0.0.16.0.20070228232320.01e13520@highstat.com> At 20:29 28/02/2007, Koen-Alonso, Mariano wrote: >Hi Alain, Hi everybody, >I am a brand-new user of Brodgar, and my apologies in advance if my >question is trivial. >I am still exploring the software and digging into the manual and >online examples, but found something that looks odd. Koen, I am not sure but it could be the case that the owner of the data gave me two years of extra data at a later stage, and that we uploaded those data. The DFA is also applied on the same data in Chapter 17 of Zuur, Ieno and Smith (2007), Analysing Ecological data (available in about 3-4 weeks). The full data are analysed in chapter 17, and give: V Diagonal V Non-diagonal Explanatory variables AIC Number of trends Explanatory variables AIC - 850.951 1 - 778.537 - 755.427 2 - 769.933 - 749.387 3 - 764.764 - 757.197 4 - 773.499 NAO 858.911 1 NAO 778.139 NAO 757.998 2 NAO 770.916 NAO 750.483 3 NAO 764.483 NAO 757.458 4 NAO 770.519 I have attached the full data, as used in the book, in a separate email. Give a shout if you don't manage to get it sorted. Alain Dr. Alain F. Zuur First author of: Analysing Ecological Data (2007). Zuur, AF, Ieno, EN and Smith, GM. Springer. 680 p. URL: www.springer.com/0-387-45967-7 Analysing Ecological data using GLMM and GAMM in R. (2008). Zuur, AF, Ieno, EN, Walker, N and Smith, GM Springer. Other books: http://www.brodgar.com/books.htm Statistical consultancy, courses, data analysis and software Highland Statistics Ltd. 6 Laverock road UK - AB41 6FN Newburgh Tel: 0044 1358 788177 Email: highstat@highstat.com URL: www.highstat.com URL: www.brodgar.com