[R] r data structures

[David Winsemius]

On Aug 16, 2012, at 1:50 PM, Schumacher, Jay S wrote:

>
> are these correct/accurate/sensible statements:
>
>  a vector is a one dimensional object.
>  a matrix is a two dimensional object.
>
>  a list is a one dimensional object.
>
> i'm working from this web page:    http://www.agr.kuleuven.ac.be/vakken/statisticsbyR/someDataStructures.htm

   You can have as many personal representations of R structures as  
you want. If you are intent on communication with other useRs, I think  
you should restrict the use of "dimension" to objects that return a  
non-NULL value from dim(). And I think one should use 'length' for the  
dimension-like aspect for atomic vectors and lists. (I originally used  
the word "attribute", but that is a specific function and applied to  
such structures would not return a "length"-value.) Dimensions is R  
can have other attributes such as names that often need to be accessed  
or manipulated

    One further source of confusion is length() applied to matrices or  
dataframes. It will return a value from matrices that is nrow()*ncol()  
and from dataframes (which are really lists under the hood) a value  
that is ncol(). The latter value is rather different than my naive  
expectation, which was that a dataframe's "length" would be the number  
of cases or rows.

    That webpage obviously has its own definition of dimension (and  
it's perfectly free to make up its own terms) but it is not one that  
is in accord with "An Introduction to R" or with typical R discourse.

-- 
David.
> -----------------------------------------------------------------
>
>
> On Aug 16, 2012, at 11:49 AM, Schumacher, Jay S wrote:
>
>>
>>
>> hi,
>> i'm trying to understand r data structures.  i see that vectors,
>> matrix, factors and arrays have a "dimension."
>> there seems to be no mention of dimensionality anywhere for lists
>> or dataframes.  can i consider lists and frames to be of fixed
>> dimension 2?
>
> About half of what you have deduced is wrong. Matrices, arrays, and
> dataframes do have dimensions, at least in technical R parlance,
> namely they have an attribute which can be queried with dim(). By
> definition matrices and dataframes have 2 dimensions. Arrays and
> matrices can be redimensioned, but dataframes cannot.
>
> Factors, lists, and atomic vectors do not have "dimensions", but they
> do have "lengths". An appropriately structured list (one with vectors
> all the same length) can be coerced to a dataframe with  
> as.data.frame().
>
> -- 
>
> David Winsemius, MD
> Alameda, CA, USA
>

David Winsemius, MD
Alameda, CA, USA