[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Coelurosaurian Analysis Matrix & Character List!!

Steven Mahon wrote-

> As you all know, I recently did a coelurosaur analysis
> w/ 30 taxa. I've fixed some errors, & reran it, coming
> out in the outfile w/ 3 trees, all the same except for
> *Ornitholestes*.

Nice first attempt at a cladistic analysis, but the matrix is far too
complete.  For instance, your codings for Coelurus-
0000000000 0000000000 000000011 0000000000 0000000000 0000000100 000000000
0000000000 0000000000 000000000 0000000000 000000000 0000000000 0000000111
100000000 0000000000 0000000000 0000001000 0000000000 0000000000 0000000000
0000000000 00000000 0000000000 0000000000 0000000000 0000000000
Not a single question mark?  When 92 of your characters are dental and
cranial, and no unambiguous cranial or dental remains have been described
for Coelurus?  Most (all?) other taxa are also coded for more characters
than possible, and probably not very accurately (if the neat Sereno-esque
distribution of derived states in the matrix is any indication).  For
instance, Allosaurus is coded as "0" for all the characters, but has at
least 26, 33, 53, 64, 87, 88, 89, 91, 95, 126, 127, 129, 131, 153, 157, 169,
180, 221, 223, 225, 227, 246 and 247.  Many characters are nonsensical, many
more are repeated.  Also, never have characters like "morphological
structure x often present" or "common ancestor with morphological structure
x", as the first makes no sense evolutionarily and the second assumes you
know the ancestor's morphology before the analysis is even run.  And never
have stand in characters just to make a clade, like your segnosaur and
pygostylian ones.  But aside from all the criticisms, it's good to see more
people trying to do this.  Just realize there's a good reason it's taken me
years just to get to character 203 in my analysis, and a good reason I get
thousands of most parsimonious trees.

Mickey Mortimer