pr<- glm(b$Group~ Age+Gender+BMI, data=b, + family=binomial(link = "logit")) Error in eval(family$initialize) : y值必需满足0 <= y <= 1

This error message is in Chinese and it says "y values must satisfy 0 <= y <= 1". This means that the response variable (b$Group) used in the logistic regression model (glm) needs to have values between 0 and 1. It is likely that the response variable is not coded correctly as a binary variable (0 or 1) and may have values outside the range of 0 and 1.