I am conducting a competing risk analysis using the Melanoma data from the {MASS} package but have been unable to obtain consistent results across the following three packages: tidycmprsk, timereg, and riskRegression.