A coding question

Mark Newman M.Newman at RBGE.ORG.UK
Wed Mar 26 09:23:08 CET 1997

Dear Joseph,

With respect to the question below, I don't know a clever way round the 
situation but I can think of a very simple one. Why not add two more 
states to character 115, as follows:

 #115. mesocarp <occurrence>/
        1. present/
        2. absent/
        3. not verifiable, layers separating
        4. not verifiable, layers not separating

and a dependency relationship 115,3/4:143:144. This will allow your data 
base to work properly and, once you get enough material of the three 
rare genera to score them, you can remove states 115,3 and 115,4.

I should point out that I use DELTA with PANKEY programs. Perhaps what 
I've said won't work with CONFOR/INTKEY.

Best wishes, Mark

> The fruit wall is made up of three layers: epicarp, mesocarp, and 
> endocarp. The presence or absence of these layers and their fusion or 
> separation are useful characteristics. So, we have the following three 
> characters:
> #115. mesocarp <occurrence>/
>        1. present/
>        2. absent/
> #143. endocarp <fused to or separating from epicarp and mesocarp>/
>        1. remaining fused to mesocarp and epicarp <IMPLICIT>/
>        2. separating from mesocarp/
>        3. separating with mesocarp from epicarp/
> #144. endocarp <fused to or separating from epicarp when mesocarp
>           absent>/
>        1. remaining fused to epicarp/
>        2. separating from epicarp/
> When the mesocarp is present (115,1), 143 is used for the fusion or 
> separation of the epicarp, mesocarp, and endocarp. When the mesocarp 
> is absent (115,2), 144 is used for the fusion or separation of the 
> epicarp and endocarp.
> When I established the dependencies 115,1:144 and 115,2:143, three 
> genera are reported as "Controlling character not coded. ... 115:143". 
> All three genera are in the same situation; character 115 is coded as 
> 115,U. These are very rare genera with very few fruits. To determine 
> the presence or absence of the mesocarp, a fruit valve must be cut or 
> broken to examine the layers. For these three genera, we do not have 
> valves that we can cut or break. It is obvious that the layers are not 
> separating, but we do not know if the mesocarp is present or absent.
> For the database to give the best results, the dependencies 115,1:144 
> and 115,2:143 should be in operation, but because of the three unknown 
> genera, it will not function with the dependencies in place. Does 
> anyone reading DELTA-L have a clever work around for this situation?

More information about the delta-l mailing list