Re: row weights in co-inertia

From: Stephane Dray (
Date: Tue Dec 21 2004 - 16:38:19 MET

  • Next message: Stephane Dray: "Re: un doute sur l'utilisation des modules Assignment New et Assign. Old"

    >Dear Stephane,
    >thanks very much for your quick and helpful answer. Now it is clear that
    >lw are the row weights and cw the column weights.
    >Of course, I meant ca$lw instead of ca$li, that was my mistake, sorry.
    >The manual I mentioned, is the complete ade4-package manual ("The ade-4
    >package, version 1.2-2, 21 May 2004"), page 65 description of "dudi" (page
    >52 in the older version1.1-2)

    ok we will check it, Thanks !

    >However, I still fail to do the randtest. The error message is:
    >"Error in randtest.coinertia(coinert, nrepet = 9999) :
    > Equal row weights expected"
      This message signify that row weights for the two tables are not
    identical. This is quite strange if you say that you have peform coinertia
    analysis and you can plot the results.

    Please can you send us exactly what are your commands (copy from R), and
    check what are
    the row weights (lw) of your two separate analyses ca$lw and fca$lw are

    What I suppose is that you have perform:
    - separate analyses
    - coinertia analysis
    - "something"
    - montecarlo tests

      And when you do "something", you change one of the two objects created by
    separate analyses (fca or ca). The randtest procedure, recall your separate
    analysis using the coinertia call. So it requires that all your objects
    exist and have not been changed. If you have changed the ca or fca object
    between the coinertia analysis and the montecarlo test, you will obtain an

    I do not know if my answer is clear ...
    I hope

    Stéphane DRAY

    Département des Sciences Biologiques
    Université de Montréal, C.P. 6128, succursale centre-ville
    Montréal, Québec H3C 3J7, Canada

    Tel : (514) 343-6111 poste 1233 Fax : (514) 343-2293
    E-mail :


    This archive was generated by hypermail 2b30 : Tue Dec 21 2004 - 16:38:22 MET