Re: making ncl quiet

From: Dennis Shea <shea_at_nyahnyahspammersnyahnyah>
Date: Tue, 15 Aug 2006 21:10:36 -0600 (MDT)

I too have asked for a way to turn off warning messages.
Perhaps, "ncl -W ...."

One of current developers looked at the possibility.
[FTR: He did not consider it a "feature". :-) ]

The problem is that the errors are not issued
from one place. Rather, they are embedded in many different
places within the core code. It would take considerable
time to implement a "-W" option.

---
So, how to turn the warnings off?
Basically, here is what is happening:
    [1] A function/procedure has an argument prototyped as
        (say) "float" or "double.
               function/procedure foo(arg0:float, ....)
        A user passes a variable [say, ii]  of type integer to "foo"
               x = foo(ii,...)
        NCL 'sees' that "foo" wants a variable of type "float".
        Thus. it creates a temporary variable of type float (say, fii).
        Then the integer ii values are "coerced" to type float.
        The temporary variable (fii) is then passed to "foo".
        If "foo" changes the values of this temporary variable,
        they will not be affect the original integer "ii" variable.
    [2] ?fix?  Well ... fix the offending arguments being input
        to the function/procedure. The error message says that the
        offending argument is "Argument 0" ... Since NCL counts
        start at 0  ... it is the 1st argement.
        So, manually do what NCL does automatically.
        fii = new ( dimsizes(ii), "float")
        x = foo(fii, ...)
        Now if "foo" changes "fii" the changed values will be returned.
good luck
D
> I guess this warning is generated when there is a mismatch in the variable 
> type between the main script and its subroutine.
> I would double-check all variable types in the script. Hope this helps.
>
>
>> I don't believe it can be done. I have asked this question before and 
>> gather than the developers consider this a 'feature.' I feel your pain.
>> 
>>> Hi all -
>>> I get a repetitive warning which I understand and have chosen to ignore:
>>> 
>>> warning:Argument 0 of the current function or procedure was coerced to the 
>>> appropriate type and thus will not change if the function or procedure 
>>> modifies its value
>>> 
>>> How do I get NCL to stop telling me about it (ie turn warnings off)?
----------------------------------------------------------------------
_______________________________________________
ncl-talk mailing list
ncl-talk_at_ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Tue Aug 15 2006 - 21:10:36 MDT

This archive was generated by hypermail 2.2.0 : Fri Aug 18 2006 - 10:43:09 MDT