[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: package description r-cran-diagnosismed



Martin Eberhard Schauer wrote:
> I'm afraid the description may need improvement.
> 
> # Package(s): r-cran-diagnosismed
> 
> Description: GNU R Diagnostic test accuracy evaluation for medical
> professionals
>  DiagnosisMed is a package to analyze data from diagnostic test accuracy
>  evaluating health conditions. It is being built to be used by health
>  professionals. This package is able to estimate sensitivity and
>  specificity from categorical and continuous test results including some
>  evaluations of indeterminate results, or compare different categorical
>  tests, and estimate reasonble cut-offs of tests and display it in a way
>  commonly used by health professionals. No graphical interface is
>  avalible yet.
> 
> -------------
> 
>  "to analyze data from diagnostic test accuracy evaluating health conditions"
>  is hard to understand (and translate). Perhaps the words are shuffled here and
>  "to analyze the accuracy of data from diagnostic tests evaluating health conditions"
>  was meant.

The original looks like word salad to me; I can only hope you've
guessed what it was intended to mean.  For a start, is it appropriate
to talk about the *data* being checked for accuracy?  The way it goes
on to talk about sensitivity and specificity makes it sound as if it's
evaluating the methodology instead.  But we really need feedback from
somebody who understands this package a lot better than I do.

[...]
> Description: GNU R Diagnostic test accuracy evaluation for medical professionals

s/Diagnostic/diagnostic/, and do we need to say that it's GNU R when
that's already obvious from the name?  Plus, this Debian package is
not literally an evaluation; it's a CRAN package, which means it's
basically a set of libraries, right?  Maybe it should be something
like this (though it's a bit of a nounpile):

  Description: medical diagnostic test accuracy analysis toolkit

>  DiagnosisMed is a package to analyze the accuracy of data from

Or "...is a GNU R package..." if that's demoted from the synopsis.

>  diagnostic tests evaluating health conditions. It was designed to be used
>  by health professionals. This package can estimate sensitivity and
>  specificity from categorical and continuous test results (including some
>  evaluations of indeterminate results), compare different categorical
>  tests, and estimate reasonble cut-offs of tests and display them in a way
>  commonly used by health professionals. No graphical interface is
>  avalible yet.

Two leftover typos: s/reasonble/reasonable/, s/avalible/available/.
But that's much more readable.
-- 
JBR	with qualifications in linguistics, experience as a Debian
	sysadmin, and probably no clue about this particular package


Reply to: