What if the biomarker high values are related to no disease? I multiplied by -1 the cutoff column of the return value of ipd2diag function. However, i cannot use log transformation like this, right? What should I do? If i set log.transformation to false the calculations are made, but with negative numbers in the yougen plot. Also, i'm not sure if the result is reliable as the distribution is logistic but the value are not loged transformated
Thanks for your tutorial. I just have a question. How can we find out what model is appropriate for our analysis? "DICS" , "CICS" or ...?
What if the biomarker high values are related to no disease? I multiplied by -1 the cutoff column of the return value of ipd2diag function. However, i cannot use log transformation like this, right? What should I do? If i set log.transformation to false the calculations are made, but with negative numbers in the yougen plot. Also, i'm not sure if the result is reliable as the distribution is logistic but the value are not loged transformated