nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Standardized residuals from Parametric Survival Fit Platform

MHz

Hi everyone,

I would like to know more about the residuals from the Parametric Survival Fit platform.

If I use this platform and plot the residuals (Residual Probability Plot), I get a plot with Standardized Residual in the X-axis and the probability in the Y-axis.

Does anyone know the formula of the Standardized Residual?

Is it the Cox-Snell Residual?

Thanks,

Accepted answers

peng_liu

For location-scale distributions (i.e. Normal, Logistic, SEV, LEV), the formula is (Y - mu(X))/sigma(X).

For log-location-scale distributions (i.e. Lognormal, Loglogistic, Weibull, Frechet, Exponential), the formula is Exp((Log(Y) - mu(X))/sigma(X)).

In both formulas, mu(X) and sigma(X) are linear functions of the covariates, depending on how your model is set up.

It is a type of Cox-Snell residual.

All comments

txnelson

I suggest that you use the "?" tool from the toolbar, and click on the plot to get the help on the plot. It may provide you with the statistical details

MHz

Hi Txnelson,

Thanks for your idea, I didn't know this way of doing (I'll keep in mind). Unfortunately, the help page does not contain information on my problem.

Thanks anyway,

peng_liu

For location-scale distributions (i.e. Normal, Logistic, SEV, LEV), the formula is (Y - mu(X))/sigma(X).

For log-location-scale distributions (i.e. Lognormal, Loglogistic, Weibull, Frechet, Exponential), the formula is Exp((Log(Y) - mu(X))/sigma(X)).

In both formulas, mu(X) and sigma(X) are linear functions of the covariates, depending on how your model is set up.

It is a type of Cox-Snell residual.

MHz

Thanks!

Regards,

Audrey_Shull

Thank you for pointing this out. Our technical writers have taken steps to add this particular information to our next iteration of the Help and documentation, which will be available this summer. Regards,

Audrey