Closed WilCrofter closed 10 years ago
Thank you very much for letting us know. We are working on repairing a few bugs and this is one of them. I hope to have it resolved in the next few days.
Sincerely, Nick
On Mon, Sep 30, 2013 at 8:07 PM, WilCrofter notifications@github.comwrote:
Question 55, Open_Intro, Module1
I have stored the values of the 'mpgCity' variable for ONLY midsize vehicles in a new variable called 'mpg.midsize', which you can view by typing 'mpg.midsize' below. Try it now.
ANSWER: mpg.midsize
Not a valid input!
An error was detected. Please notify the administator.
exists("mpg.midsize") [1] FALSE
data("mpg.midsize") Warning message: In data("mpg.midsize") : data set ‘mpg.midsize’ not found
Looks like the subset of cars is not in the package or dta files. Nor is it referred to in mod.info:
mod.info Instructor.name. Nicholas.Carchedi 1 Course name: Open Intro 2 Module name: Introduction to Data 3 Packages required for this module (separated by commas): openintro, plotrix 4 Data sets required for this module (separated by commas): cars
— Reply to this email directly or view it on GitHubhttps://github.com/ncarchedi/swirl/issues/5 .
Nicholas A. Carchedi ScM Student, Biostatistics Johns Hopkins Bloomberg School of Public Health 301.802.0573 | nick.carchedi@gmail.com
You are absolutely correct in your assessment of the problem above. What I meant to do is this: Have swirl pre-define variables for the user according the instructor's wishes. It seems that the best place to do this would be the mod.info
data frame where the rest of the "meta-data" regarding the module is stored. This is where we tell swirl what packages and data sets are required for each module.
We could simply add another entry for variable definitions that would be parsed and executed upon beginning the module. This would not be hard to do, but I'd have to make a few adjustments. Let me know if you agree this is a good (at least temporary) solution.
Nick
I agree. Least chance of side effects,
etc.
As it is, the mod contains code to define mpg.midsize at line 28:
mod$Correct.Answer[28] [1] "mpg.midsize <- cars$mpgCity[cars$type==\"midsize\"]"
but this code is not executed the first time it's encountered, I
think because line 28's Output.Type is text. The entry might be
left over from an earlier version of the mod.
Anyway, the entry is parsed and eval'd if you resume after
aborting first time through. I'm guessing, but that might be
because, upon resuming, all correct answers containing '<-'
which the user has already covered are executed. Maybe execution
should be limited to correct answers with output type question for
consistency.
Cheers,
Bill
On 10/1/2013 10:38 PM, Nick Carchedi wrote:
You are absolutely correct in your assessment of the problem
above. What I meant to do is this: Have swirl pre-define
variables for the user according the instructor's wishes. It
seems that the best place to do this would be the mod.info
data frame where the rest of the "meta-data" regarding the
module is stored. This is where we tell swirl what packages and
data sets are required for each module.
We could simply add another entry for variable definitions that
would be parsed and executed upon beginning the module. This
would not be hard to do, but I'd have to make a few adjustments.
Let me know if you agree this is a good (at least temporary)
solution.
Nick
—
Reply to this email directly or view
it on GitHub.
Noted. Give me a few days on this and I will implement a solution.
Thanks again, Nick On Oct 2, 2013 8:47 AM, "WilCrofter" notifications@github.com wrote:
I agree. Least chance of side effects, etc. As it is, the mod contains code to define mpg.midsize at line 28:
mod$Correct.Answer[28] [1] "mpg.midsize <- cars$mpgCity[cars$type==\"midsize\"]"
but this code is not executed the first time it's encountered, I think because line 28's Output.Type is text. The entry might be left over from an earlier version of the mod. Anyway, the entry is parsed and eval'd if you resume after aborting first time through. I'm guessing, but that might be because, upon resuming, all correct answers containing '<-' which the user has already covered are executed. Maybe execution should be limited to correct answers with output type question for consistency. Cheers, Bill On 10/1/2013 10:38 PM, Nick Carchedi wrote:
You are absolutely correct in your assessment of the problem above. What I meant to do is this: Have swirl pre-define variables for the user according the instructor's wishes. It seems that the best place to do this would be the mod.info data frame where the rest of the "meta-data" regarding the module is stored. This is where we tell swirl what packages and data sets are required for each module. We could simply add another entry for variable definitions that would be parsed and executed upon beginning the module. This would not be hard to do, but I'd have to make a few adjustments. Let me know if you agree this is a good (at least temporary) solution. Nick — Reply to this email directly or view it on GitHub.
— Reply to this email directly or view it on GitHubhttps://github.com/ncarchedi/swirl/issues/5#issuecomment-25535314 .
Nick,
Sorry for this afterthought. I know you are swamped, 5th week of
semester or so.
But what do you think of adding another type, 'source', which
worked a bit like 'figure' but merely sourced a R file? That way
the timing of its execution could be controlled, as it is with
figure.
Bill
On 10/2/2013 11:22 AM, Nick Carchedi wrote:
Noted. Give me a few days on this and I will implement
a solution.
Thanks again,
Nick
On Oct 2, 2013 8:47 AM, "WilCrofter"
<notifications@github.com> wrote:
>
> I agree. Least chance of side effects,
> etc.
> As it is, the mod contains code to define mpg.midsize at line
28:
>
> > mod$Correct.Answer[28]
> [1] "mpg.midsize <- cars$mpgCity[cars$type==\"midsize\"]"
>
> but this code is not executed the first time it's
encountered, I
> think because line 28's Output.Type is text. The entry might
be
> left over from an earlier version of the mod.
> Anyway, the entry is parsed and eval'd if you resume after
> aborting first time through. I'm guessing, but that might be
> because, upon resuming, all correct answers containing
'<-'
> which the user has already covered are executed. Maybe
execution
> should be limited to correct answers with output type
question for
> consistency.
> Cheers,
> Bill
> On 10/1/2013 10:38 PM, Nick Carchedi wrote:
>
> You are absolutely correct in your assessment of the problem
> above. What I meant to do is this: Have swirl pre-define
> variables for the user according the instructor's wishes. It
> seems that the best place to do this would be the mod.info
> data frame where the rest of the "meta-data" regarding the
> module is stored. This is where we tell swirl what packages
and
> data sets are required for each module.
> We could simply add another entry for variable definitions
that
> would be parsed and executed upon beginning the module. This
> would not be hard to do, but I'd have to make a few
adjustments.
> Let me know if you agree this is a good (at least temporary)
> solution.
> Nick
> —
> Reply to this email directly or view
> it on GitHub.
>
> —
> Reply to this email directly or view it on
GitHubhttps://github.com/ncarchedi/swirl/issues/5#issuecomment-25535314
.
>
—
Reply to this email directly or view
it on GitHub.
I think this is a great idea. I'll have to deal with this one on my own, since I'll modify the excel templates sent to instructors in the first place. I don't include the templates in the package, just the .rda files that are derived from them.
Thanks again, Nick
On Wed, Oct 2, 2013 at 1:08 PM, William R Bauer notifications@github.comwrote:
Nick, Sorry for this afterthought. I know you are swamped, 5th week of semester or so. But what do you think of adding another type, 'source', which worked a bit like 'figure' but merely sourced a R file? That way the timing of its execution could be controlled, as it is with figure. Bill
On 10/2/2013 11:22 AM, Nick Carchedi wrote: Noted. Give me a few days on this and I will implement a solution.
Thanks again,
Nick
On Oct 2, 2013 8:47 AM, "WilCrofter" notifications@github.com wrote:
I agree. Least chance of side effects,
etc.
As it is, the mod contains code to define mpg.midsize at line 28:
mod$Correct.Answer[28]
[1] "mpg.midsize <- cars$mpgCity[cars$type==\"midsize\"]"
but this code is not executed the first time it's encountered, I
think because line 28's Output.Type is text. The entry might be
left over from an earlier version of the mod.
Anyway, the entry is parsed and eval'd if you resume after
aborting first time through. I'm guessing, but that might be
because, upon resuming, all correct answers containing '<-'
which the user has already covered are executed. Maybe execution
should be limited to correct answers with output type question for
consistency.
Cheers,
Bill
On 10/1/2013 10:38 PM, Nick Carchedi wrote:
You are absolutely correct in your assessment of the problem
above. What I meant to do is this: Have swirl pre-define
variables for the user according the instructor's wishes. It
seems that the best place to do this would be the mod.info
data frame where the rest of the "meta-data" regarding the
module is stored. This is where we tell swirl what packages and
data sets are required for each module.
We could simply add another entry for variable definitions that
would be parsed and executed upon beginning the module. This
would not be hard to do, but I'd have to make a few adjustments.
Let me know if you agree this is a good (at least temporary)
solution.
Nick
—
Reply to this email directly or view
it on GitHub.
—
Reply to this email directly or view it on GitHubhttps://github.com/ncarchedi/swirl/issues/5#issuecomment-25535314
.
— Reply to this email directly or view it on GitHub.
— Reply to this email directly or view it on GitHubhttps://github.com/ncarchedi/swirl/issues/5#issuecomment-25557004 .
Nicholas A. Carchedi ScM Student, Biostatistics Johns Hopkins Bloomberg School of Public Health 301.802.0573 | nick.carchedi@gmail.com
This should no longer be a problem. I just pushed a temporary patch that will have to do until we find a more elegant solution.
Question 55, Open_Intro, Module1
I have stored the values of the 'mpgCity' variable for ONLY midsize vehicles in a new variable called 'mpg.midsize', which you can view by typing 'mpg.midsize' below. Try it now.
ANSWER: mpg.midsize
Not a valid input!
An error was detected. Please notify the administator.
> exists("mpg.midsize") [1] FALSE
> data("mpg.midsize") Warning message: In data("mpg.midsize") : data set ‘mpg.midsize’ not found
Looks like the subset of cars is not in the package or dta files. Nor is it referred to in mod.info:
> mod.info Instructor.name. Nicholas.Carchedi 1 Course name: Open Intro 2 Module name: Introduction to Data 3 Packages required for this module (separated by commas): openintro, plotrix 4 Data sets required for this module (separated by commas): cars
If you resume after the abort, mpg.midsize exists. The reason appears to be that
mod[23, "Correct.Answer"]
, which definesmpg.midsize
was executed upon resumption along with other Correct.Answer code involving assignment. This entry was the "correct answer" to a "text" type Output which did not require an answer, hence was not executed when first encountered.