[ESS] Does R fix() or edit() work for you these days?]

[8rino-Luca Pantani]

Yes, Martin you're right.
May be that "rules" is a too stringent word.
I would say that I was asking for "good habits" or "tricks" or 
"self-discipline",
rather than strict rules.
Most of all, when you get back to old analyses, when you do not realize 
promptly what you did.
For example at the beginning I didn't use the form lm.<topic> but only 
<topic>
ending in a overpopulated data region, in which it was almost impossible 
to recognize what was what. Especially after a nice week-end ;-)


Martin Maechler ha scritto:
> yes.
>
> OTOH, 8rino asks about "your favorite way"  which is much less
> stringent than "rules" and rather a matter of preferences.
>
> My personal taste / opinion is to *not* use long variable names,
> and I do work with "." here, where other people would prefer
> other letters/styles.
> E.g., I typicall call most of my data frames  
>   'd.<topic>'
> and then, the models I fit to this data, something like
>   'lm1.<topic>'
>   'lm2.<topic>'
>   'glm1.<topic>'
> etc; but all this depends on the circumstances and I'd always
> "allow myself" to differ from such a scheme whenever I see
> reasons.
>
> One good reason for the above scheme is that you want names that
> differ rather in the beginning than in the end, so
>  <topic><subtopic>model1
>  <topic><subtopic>model2
>  <topic><subtopic>model3
> seems less sensible to me -- both for reading and for writing
> such code.
>
>   
Another good reason is that
ls()
give all the dataframes separated from models, vectors and so on.