I downloaded the TCGA prostate adenocarcinoma clinical data from the following url:
In the clinicalMatrix file, the "biochemical_recurrence" field for the following sample ids are all "NO", presumably indicating that these samples did not have biochemical recurrence. However, their "days_to_first_biochemical_recurrence" are not 0. Could somebody please explain if this is an error? Should I assume that because the "days_to_first_biochemical_recurrence" exists the sample actually had BCR?
For other sample ids where the "biochemical_recurrence" is "NO", the "days_to_first_biochemical_recurrence" is absent or NaN.
Also, the following sample ids' "biochemical_recurrence" field is NaN, but the "days_to_first_biochemical_recurrence" is not absent or NaN. What should I do for these samples?