[BUG]nauc() function about drda

[albertopessia]

In order to debug the code I need to know first the parameters you used.

What are the values of xlim, ylim, and the model coefficients (you can extract them with coef(your_drda_fit_object))? Is the model a 4-parameter logistic function?

[ofleitas]

Hello I plotted with the default parameters. Looking at the graph I think that xlim is (-2,2) (this is log(dose)) and the ylim (0,100) (This is in percent). The model is a 4-parameter logistic function and the coefficient are alpha beta eta phi 9.4002937 99.9950666 4.7709601 -0.7728488

…

On Wed, Mar 16, 2022 at 1:47 AM Alberto Pessia ***@***.***> wrote: In order to debug the code I need to know first the parameters you used. What are the values of xlim, ylim, and the model coefficients (you can extract them with coef(your_drda_fit_object))? Is the model a 4-parameter logistic function? — Reply to this email directly, view it on GitHub <#4 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AM624NAXTMHESJBBCPG6NMLVAGGZBANCNFSM5QRFIKWA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[albertopessia]

The bug was in the specification of the ylim argument. By default ylim is defined for a relative scale of c(0, 1) instead of a percentage of c(0, 100).

Since the curve in this example is defined for percentages with values in c(9.4, 99.995), it does not intersect in this particular case the default range c(0, 1). Therefore, inner computations failed with a logarithm of a negative number.

I now fixed the bug by adding extra checks in the latest 2.0 version. CRAN version is coming soon but it is possible to install the development version through GitHub (see README in the homepage).

However, if your data represent percentages, remember to set ylim = c(0, 100) when calling nauc() and naac().

Note: New 2.0 version has a slightly different parameterization from before. Instead of a lower bound (alpha) and upper bound (beta) parameters, we now use the same parameterization of the Emax model, that is the value at -Inf (alpha) and the maximum effect at Inf (delta) (check help(drda) to know more).