[R] R-specific Software Requirement Specification

[Doran, Harold]

This is in fact a *very* good suggestion, Duncan. One could easily rewrite any existing R function by looking at the help page I believe. So, I could in fact begin by writing the help page which we know details the function inputs, details about each argument, what the function is expected to output etc. 

I suppose it is easy enough to add an "inherits from" or outputs an object of class "xyz" to this for my purpose.

 

-----Original Message-----
From: Duncan Murdoch [mailto:murdoch.duncan at gmail.com] 
Sent: Wednesday, September 07, 2016 12:46 PM
To: Doran, Harold <HDoran at air.org>; r-help at r-project.org
Subject: Re: [R] R-specific Software Requirement Specification

On 07/09/2016 11:35 AM, Doran, Harold wrote:
> I'm building a large program with many different people contributing to the coding in R and so it needs a well-articulated design spec. The program will have many different functions that must interact with each other, but the individual functions will be written by different people.
>
> I'm curious if anyone has an R-specific SRS document to share that they have used for a similar purpose listing the objectives for each function, class definition, generics, what the function inherits from, and so on, or perhaps even a useful template for such work.

The Rd help pages do some of this.  They aren't so good at describing the class hierarchy but are good at specifying individual functions.

Duncan Murdoch