[R] Error when using qvalue function

[Ana Marija]

Hello,

I am am correct to assume that here:

pvals=q$META
qval_obj=qvalue(pvals) #is false discovery rate
pi1=1-qval_obj$pi0 #TPR
pi1  #TPR

pi1 represents True Positive Rate?

Thanks
Ana

On Mon, Oct 28, 2019 at 9:36 PM William Dunlap <wdunlap using tibco.com> wrote:
>
> With qvalue-2.16.0 and R-3.6.1 many calls to qvalue::qvalue die in its call to qvalue::pi0est.  E.g.,
>
> > trace(smooth.spline, quote(cat("y=", deparse(y), "\n")))
> Tracing function "smooth.spline" in package "stats"
> [1] "smooth.spline"
> > qvalue::pi0est(c(.01, .01, .06))
> Tracing smooth.spline(lambda, pi0, df = smooth.df) on entry
> y= c(NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_,  NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_,  NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_)
> Error in smooth.spline(lambda, pi0, df = smooth.df) :
>   missing or infinite values in inputs are not allowed
>
> You should submit a bug report to
> > packageDescription("qvalue")$URL
> [1] "http://github.com/jdstorey/qvalue"
>
> Bill Dunlap
> TIBCO Software
> wdunlap tibco.com
>
>
> On Mon, Oct 28, 2019 at 3:13 PM Jim Lemon <drjimlemon using gmail.com> wrote:
>>
>> Hi Ana,
>> Seems to work without error for me:
>>
>> # installed qvalue_1.26.0 from CRAN archive
>> library(qvalue)
>>  pvals<-c(6.919239e-02,1.073784e-01,1.218613e-01,1.586202e-01,
>>  1.370340e-01,3.452574e-02,2.545619e-01,1.676715e-02,8.571197e-01,
>>  8.649025e-01,1.777414e-02,6.801867e-01,6.873085e-01,
>>  1.276566e-01,5.907002e-02,2.343207e-02,2.078125e-02,6.404511e-02,
>>  3.306593e-02,9.411259e-04,1.038989e-03,4.734674e-05,
>>  3.422489e-05,5.606264e-05,6.322817e-02,1.291268e-02,3.452596e-03,
>>  1.770753e-01,3.285821e-01,2.292055e-02,5.503168e-02,
>>  6.940048e-01,3.638889e-03,6.799640e-01,1.301045e-02,6.794010e-01,
>>  2.339327e-02,1.038529e-01,3.137687e-04,2.148050e-02,
>>  1.783068e-01,1.764518e-01,6.386686e-03,1.670062e-02,3.220291e-01,
>>  5.568613e-01,8.886102e-01,5.031040e-01,3.760079e-01,
>>  6.638034e-02,9.419648e-03,5.885266e-03,1.539809e-02,5.296551e-03,
>>  2.425230e-02,5.023091e-01,4.547284e-03,1.850796e-01,
>>  9.389289e-02,6.544873e-03,3.031956e-03,7.772671e-03,9.073974e-03,
>>  9.118352e-02,4.075408e-04,6.902206e-01,6.929767e-02,
>>  1.897121e-02,6.693074e-02,1.111308e-02,1.286147e-02,4.515834e-02,
>>  8.886941e-01,8.891051e-01,3.792846e-01,5.368898e-01,
>>  2.323894e-01,3.220141e-01,7.320883e-02,9.642521e-03,6.024415e-01,
>>  2.459322e-02,2.873351e-01,8.477168e-01,1.351068e-02,
>>  1.053550e-01,4.812686e-01,1.404957e-01,9.835912e-02,4.373995e-01,
>>  8.803856e-02)
>> qval_obj=qvalue(pvals)
>> qval_obj$pi0
>> [1] 0.1981095
>> pi1=1-qval_obj$pi0
>> pi1
>> [1] 0.8018905
>>
>> Jiim
>>
>> On Tue, Oct 29, 2019 at 8:45 AM Ana Marija <sokovic.anamarija using gmail.com> wrote:
>> >
>> > Hello,
>> >
>> > I am trying to calculate True Positive Rate, TPR with this procedure:
>> >
>> >     pvals=q$METAge
>> >     qval_obj=qvalue(pvals) #is false discovery rate
>> >     pi1=1-qval_obj$pi0 #TPR
>> >     pi1  #TPR
>> >
>> > But I am getting this error:
>> >
>> >     Error in smooth.spline(lambda, pi0, df = smooth.df) :
>> >        missing or infinite values in inputs are not allowed
>> >
>> > I have 91 p values and they look like this:
>>
>> ______________________________________________
>> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.