[R] warning on generic function when building R package

[Duncan Murdoch]

On 27/10/2015 3:52 PM, David Stevens wrote:
> Sorry to be a late comer to this problem but I'm having a similar 
> issue.  My function is called by ode from deSolve
> 
> ADM1_C <- function(t,state,parameters,...){
> 
> with(as.list(c(state,parameters)), {
>    # do some stuff here and return a list containing a vector of 
> derivatives to ode
> ...
> })
> }
> in which 't' is a time and 'state' and 'parameters' are numeric vectors 
> with each element named. When I invoke the solver,
> 
> times <- d.df$time
> out <- as.data.frame(ode(y = state,times = times,func = ADM1_C,parms = 
> parameters))
> 
> I get ~three Note:s for each value in 'state'  (sometimes more). The 
> calculation is successful but the issue is a little puzzling.
> Here's 'state'
> state
>       Ssu      Saa      Sfa      Sva      Sbu     Spro Sac      Sh2     
> Sch4      Sic
> 0.300000 0.001000 0.300000 0.300000 0.300000 0.300000 0.300000 0.000001 
> 0.000010 0.040000
>       Sin       Si       Xc      Xch      Xpr      Xli Xsu      Xaa      
> Xfa      Xc4
> 0.010000 0.020000 0.300000 0.026000 0.300000 0.030000 0.400000 1.100000 
> 0.200000 0.410000
>      Xpro      Xac      Xh2       Xi  Scation   Sanion Sva_m    Sbu_m   
> Spro_m    Sac_m
> 0.137000 0.700000 0.010000 5.000000 0.040000 0.020000 0.090000 0.090000 
> 0.074000 0.169000
>   Shco3_m     Snh3  Sgas_h2 Sgas_ch4 Sgas_co2
> 0.009700 0.016000 0.020000 0.020000 0.037600
> 
> The messages are something like this ...
> Note: no visible binding for global variable 'Sin'
> Note: no visible binding for global variable 'Snh3'
> Note: no visible binding for global variable 'Scation'
> Note: no visible binding for global variable 'Shco3_m'
> .... (~130 rows of these messages)

I don't see how this has anything to do with the previous problem.
> 
> Neither Duncan's nor Jim's solution solved the problem.  I can't change 
> (at least I don't think so) how ode(...) calls the function except via 
> the documentation. Any ideas? Does it matter? Could it be ode strips out 
> the names, but only sometimes?

You will need to post something that we can reproduce if you want help
with it.  (Please do this on a new thread, since you really have nothing
to do with generic functions as far as I can see.)

Or you can add cat() and print() statements to your function (or set
breakpoints in a debugger) and you'll be able to see what it's getting.

Duncan Murdoch

> 
> signed Grasping at Straws (David)
> 
> David K Stevens, P.E., Ph.D.
> Professor and Head, Environmental Engineering
> Civil and Environmental Engineering
> Utah Water Research Laboratory
> 8200 Old Main Hill
> Logan, UT  84322-8200
> 435 797 3229 - voice
> 435 797 1363 - fax
> david.stevens at usu.edu
> 
> 
> On 10/19/2015 2:32 PM, carol white via R-help wrote:
>> In effect, this works
>> but whether I use x or x.init, y or y.init in plot.func, I get
>>   
>> no visible binding for global variable ‘x.init’no visible binding for global variable ‘y.init’
>> Regards,
>>
>>       On Monday, October 19, 2015 9:59 PM, Duncan Murdoch <murdoch.duncan at gmail.com> wrote:
>>     
>>
>>   On 19/10/2015 3:50 PM, carol white wrote:
>>> Thanks Murdoch.
>>>
>>> defining
>>> plot.func<- function(x=x.init, y=y.init, arg3, arg4, "title", col, arg5)
>>>
>>> and if plot doesn't take the exact parameters of plot.func but modified
>>> of these parameters
>>> plot(x=x.pt,y=y.pt,xlim = c(0, 10), ylim = c(0,1), xlab= "xlab",
>>> ylab="ylab", main = "title", col = col,type = "l")
>>>
>>> then, how to define and invoke to be consisent?
>> I don't really understand your question, but this is all about the
>> function header for plot.func, not the call you make to plot().  You
>> need to name the first argument as "x", you need to include "..." as an
>> open argument, and you need a legal header.  So this would be okay:
>>
>>
>> plot.func<- function(x=x.init, y=y.init, arg3, arg4,
>>                      main = "title", # can't skip the arg name
>>                      col, arg5,
>>                      ...)  {          # can't skip the dots
>>
>> Duncan Murdoch
>>
>>> Regards,
>>>
>>> On Monday, October 19, 2015 7:45 PM, Duncan Murdoch
>>> <murdoch.duncan at gmail.com> wrote:
>>>
>>>
>>> On 19/10/2015 1:29 PM, carol white via R-help wrote:
>>>
>>>> Hi,I have invoked plot in a function (plot.func) as follows but when I
>>> check the built package, I get a warning:
>>>> plot(x.pt,y.pt,xlim = c(0, 10), ylim = c(0,1), xlab= "xlab",
>>> ylab="ylab", main = "title", col = col,type = "l")
>>>> R CMD check my.package
>>>> checking S3 generic/method consistency ... WARNING
>>>> plot:
>>>>    function(x, ...)
>>>> plot.func:
>>>>    function(x.pt, y.pt, arg3, arg4, "title", col, arg5)
>>>>
>>>> See section ‘Generic functions and methods’ in the ‘Writing R
>>>> Extensions’ manual.
>>>> Which plot argument is illegitimate or missing and how to eliminate
>>> the warning?
>>>
>>>
>>> The first argument to plot.func needs to be called "x" if you want to
>>> use it as a method.  Method signatures need to be consistent with the
>>> generic signature.
>>>
>>> Duncan Murdoch
>>>
>>>
>>>
>>>
>>
>>
>>    
>> 	[[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
> 
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>