R help - Oct 2009 - tapply with multiple arguments that are not part of the same data frame

Hi all,

I would like to invoke a function that takes multiple arguments (some of
which are specified columns in the data frame, and others that are
independent of the data frame) on split parts of a data frame, how do I do
this?

For example, let's say I have a data frame
>fitness_data
name  height  weight  country
rob      5.8        200      usa
nancy  5.5        140      germany
jen       5.6        150      usa
clark     5.10      210     germany
matt     5.9         280     canada
ralph    6           270     canada
...
...

Now let us say I have a function,  my_func(h, w, noise, dir), which takes as
input:
(1) a vector of heights
(2) a vector of weights
(3) a user-input numeric "noise" value
(4) a user-input string "dir" for the directory to output the end
result of
the function to

This function does some calculations on the input data and outputs a
dataframe that is then written to a file in the "dir" directory.

If I want to apply this function to data grouped by each country in the
"fitness_data" dataframe, how would I do this? I tried looking through
the
mailing archives, but couldn't nail down the solution. I tried something
like

split(mapply( function(a,b,c,d) my_func(fitness_data$h, fitness_data$w, 2.5,
my_directory)), fitness_data$country)

but this considered fitness_data$h, and fitness_data$w in each single row
for a country, rather than a vector of heights or weights across all rows
corresponding to that country.

Thanks!

	[[alternative HTML version deleted]]

I just realized my earlier post of my question below was not in
"Plain" Text mode, hence the repeat post...apologies!
Kavitha


On Thu, Oct 22, 2009 at 4:19 PM, Kavitha Venkatesan
<kavitha.venkatesan at gmail.com> wrote:
> Hi all,
>
> I would like to invoke a function that takes multiple arguments (some of
> which are specified columns in the data frame, and others that are
> independent of the data frame) on split parts of a data frame, how do I do
> this?
>
> For example, let's say I have a data frame
>>fitness_data
> name? height? weight? country
> rob????? 5.8??????? 200????? usa
> nancy? 5.5??????? 140????? germany
> jen?????? 5.6??????? 150????? usa
> clark???? 5.10????? 210???? germany
> matt???? 5.9???????? 280???? canada
> ralph??? 6?????????? 270???? canada
> ...
> ...
>
> Now let us say I have a function,? my_func(h, w, noise, dir), which takes
as
> input:
> (1) a vector of heights
> (2) a vector of weights
> (3) a user-input numeric "noise" value
> (4) a user-input string "dir" for the directory to output the end
result of
> the function to
>
> This function does some calculations on the input data and outputs a
> dataframe that is then written to a file in the "dir" directory.
>
> If I want to apply this function to data grouped by each country in the
> "fitness_data" dataframe, how would I do this? I tried looking
through the
> mailing archives, but couldn't nail down the solution. I tried
something
> like
>
> split(mapply( function(a,b,c,d) my_func(fitness_data$h, fitness_data$w,
2.5,
> my_directory)), fitness_data$country)
>
> but this considered fitness_data$h, and fitness_data$w in each single row
> for a country, rather than a vector of heights or weights across all rows
> corresponding to that country.
>
> Thanks!
>
>