search

lgessl / patroklos

An R package pipelining binary classification of survival
https://lgessl.github.io/patroklos/
Other
0 stars 0 forks source link

Too many benchmark points in comparative 2D plot #9

Closed lgessl closed 8 months ago

lgessl commented 8 months ago

There must be duplicates in the data points plotted for the benchmark in the comparative performance plot at the end of assessment_center().

lgessl commented 8 months ago

Fixed the issue. Problem was that prepare_and_predict() prepared the actual risk according to model_spec$time_cutoffs—and not perf_plot_spec$pivot_time_cutoff().