Re: strange behavior

Yes, I see the problem.

NCL uses "pass-by-reference" for arguments to functions
and procedures. Hence, if you change the value of an argument
inside a function or procedure
*the changed value will be returned ... just like fortran!*

For NCL's *built-in* functions, we *never* change the value of
an input argument. I recommend that all users do the same thing.

If necessary, use temporary variables within a function to
preserve the original values.

Cheers
D

----
function TEST (t)
begin
    t  =  t - 273.15    ; change the argument's value
    tc = t                  ; tc AND t will change
    return (tc)
 end
begin
    T   = 300.
    TT = TEST( T )                           ; T keeps changing
    print("T="+T+"   TT="+TT)  
    print("T="+T+"   TT="+TT)
    
    
Mateus da Silva Teixeira wrote:
>> Mateus da Silva Teixeira wrote:
>>     
>>> Dear NCL users,
>>>
>>> I'm having a strange behavior of mixhum_ptrh function and my
>>> user-defined function. Both functions give the specific humidity given
>>> temperature, pressure and relative humidity. I printed their results to
>>> compare them, but I'm getting this:
>>>
>>> ncl 1> p  = 1000.
>>> ncl 2> tk =   18. + 273.15
>>> ncl 3> rh =   46.5
>>> ncl 4> print(qSpec(tk,rh,p)+"   "+mixhum_ptrh (p, tk, rh, 2))
>>> (0)     5.96583e-05   -0.00771059
>>> ncl 5> print(qSpec(tk,rh,p)+"   "+mixhum_ptrh (p, tk, rh, 2))
>>> (0)     nan   -7.65217e-05
>>> ncl 6> print(qSpec(tk,rh,p)+"   "+mixhum_ptrh (p, tk, rh, 2))
>>> (0)     -1.73974   -7.72574e-07
>>> ncl 7> print(qSpec(tk,rh,p)+"   "+mixhum_ptrh (p, tk, rh, 2))
>>> (0)     1.86416e-06   8.26161e-10
>>>
>>> As you can see, the values isn't the same in successive calls.
>>> I'm using NCL 4.3.0.
>>>
>>> Do someone already seen this?
>>>
>>> Thanks,
>>>
>>> Mateus
>>>   
>>>       
>>     
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFGgWbitBm/Y0ahARYRAl5zAJ4z23KzCFd3djp/xhDDHnVhKdD7QQCfQS39
> 8esddxboiW2JnyZsnJE2D38=
> =H/Bd
> -----END PGP SIGNATURE-----
>   
> ------------------------------------------------------------------------
>
> undef("qSpec")
> ; function that returns specific humidity given
> ; air temperature (Kelvin), relative humidity (%)
> ; and pressure (hPa).
> ;
> ; Mateus Teixeira, jun/2007
> function qSpec(T:float,RH:float,p:float)
> local c0,A,B,epsilon,es,e,r,q
> begin
>   ; constants used
>   ; c0, A e B were taken from (1994), pag. 117.
>   ; it's valid in the range
>   ; -35C <= T <= 35C
>   c0 = 6.112
>   A  = 17.67
>   B  = 243.5
>   epsilon = 0.622 ; Rd / Rv
>
>   ; some necessary transformations
>   T  = T-273.15 ; Kelvin -> Celsius
>   p  = p*100.0  ; hPa -> Pa
>   RH = RH/100.0
>   
>   ; equilibrium (saturation) vapor pressure
>   es = c0*exp(A*T/(T+B))
>   ; vapor pressure
>   e  = RH*es
>   ; mix ratio
>   r  = e*epsilon/(p-e)
>   ; specific humidity
>   q  = r/(1.+r)
>   
>   return(q)
> end
>   
-- 
======================================================
Dennis J. Shea                  tel: 303-497-1361    |
P.O. Box 3000                   fax: 303-497-1333    |
Climate Analysis Section                             |
Climate & Global Dynamics Div.                       |
National Center for Atmospheric Research             |
Boulder, CO  80307                                   |
USA                        email: shea 'at' ucar.edu |
======================================================
_______________________________________________
ncl-talk mailing list
ncl-talk_at_ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk