R help - Sep 2019 - Help needed with eval parse

Hello R-users,

I have been running a script which produces objects based on the
column names of a data.frame. The column names are of the form CB_1-1,
CB_1-2, etc. Now this calculation was rather long and memory
intensive, so I would rather not have to do it again after fixing the
column names using "make.names". As a consequence, I am left with a
bunch of R objects with `-` in the name.
Accessing them is proving challenging and any help would be appreciated.

Reproducible example:
`cb_1-2` <- "hello world"
t <- "cb_1-2"
t <- as.name(t)
t <- eval(parse(text = t))

Error in eval(parse(text = t)) : object 'cb_1' not found

Thanks.
huzefa

On 18/09/2019 8:43 a.m., Huzefa Khalil wrote:
> Hello R-users,
> 
> I have been running a script which produces objects based on the
> column names of a data.frame. The column names are of the form CB_1-1,
> CB_1-2, etc. Now this calculation was rather long and memory
> intensive, so I would rather not have to do it again after fixing the
> column names using "make.names". As a consequence, I am left with
a
> bunch of R objects with `-` in the name.
> Accessing them is proving challenging and any help would be appreciated.
> 
> Reproducible example:
> `cb_1-2` <- "hello world"
> t <- "cb_1-2"
> t <- as.name(t)
> t <- eval(parse(text = t))
> 
> Error in eval(parse(text = t)) : object 'cb_1' not found
After t <- as.name(t), you already have language:  no need to parse it 
again.  So

   eval(t)

works.  If you have more complicated expressions, use call() to set them 
?p.  For example, call("paste0", t, "!") evaluates to

   paste0(`cb_1-2`, "!")

and evaluating that expression via

   eval(call("paste0", t, "!"))

gives

[1] "hello world!"

Don't go back and forth between language objects and text 
representations of them, because it's hard to do that without 
introducing changes.  In other words, don't use eval(parse()).

Duncan Murdoch

That worked! Thanks for the explanation.

On Wed, Sep 18, 2019 at 10:06 AM Duncan Murdoch
<murdoch.duncan at gmail.com> wrote:
>
> On 18/09/2019 8:43 a.m., Huzefa Khalil wrote:
> > Hello R-users,
> >
> > I have been running a script which produces objects based on the
> > column names of a data.frame. The column names are of the form CB_1-1,
> > CB_1-2, etc. Now this calculation was rather long and memory
> > intensive, so I would rather not have to do it again after fixing the
> > column names using "make.names". As a consequence, I am left
with a
> > bunch of R objects with `-` in the name.
> > Accessing them is proving challenging and any help would be
appreciated.
> >
> > Reproducible example:
> > `cb_1-2` <- "hello world"
> > t <- "cb_1-2"
> > t <- as.name(t)
> > t <- eval(parse(text = t))
> >
> > Error in eval(parse(text = t)) : object 'cb_1' not found
>
> After t <- as.name(t), you already have language:  no need to parse it
> again.  So
>
>    eval(t)
>
> works.  If you have more complicated expressions, use call() to set them
> ?p.  For example, call("paste0", t, "!") evaluates to
>
>    paste0(`cb_1-2`, "!")
>
> and evaluating that expression via
>
>    eval(call("paste0", t, "!"))
>
> gives
>
> [1] "hello world!"
>
> Don't go back and forth between language objects and text
> representations of them, because it's hard to do that without
> introducing changes.  In other words, don't use eval(parse()).
>
> Duncan Murdoch