R 拒绝所有变量

在R中进行逻辑回归时，所有p值均为1，拒绝了所有假设。会有什么错误

# Build Logistic Model
logdown <- glm(classification ~ bp+bgr+bu+sod+pot+hemo+pcv+wc+rc, family = "binomial", data = down_train)

---

你可以检查这个问题。嘿，你的分类栏错了。在进行逻辑回归时，你怎么能得到384.391182的截距？欢迎使用！看起来你的数据有很强的多重共线性。系数值是正确的，但标准误差很大，剩余偏差很小。一个接一个地从数据中删除变量，看看它是否在某个时候变得更好。

summary(logdown)

Call: glm(formula = classification ~ bp + bgr + bu + sod + pot + hemo
     pcv + wc + rc, family = "binomial", data = down_train) 

 Deviance Residuals: 
           Min             1Q         Median             3Q            Max  
 -0.0000231166  -0.0000000211   0.0000000000   0.0000000211   0.0000310084   

 Coefficients:
                   Estimate     Std. Error z value Pr(>|z|) 
(Intercept)     384.391182 1244660.587236       0        1 
bp                0.954292    5868.460874       0        1 
bgr               0.503592    1472.911453       0        1 
bu               -0.104507     857.081651       0        1
sod              -2.889344   10937.812900       0        1 
pot               7.801181   54746.963038       0        1 
hemo             -6.398532   25335.287408       0        1 
pcv              -1.060024    6935.366539       0        1 
wc                0.001903      25.199416       0        1 
rc               -6.043063   47852.993574       0        1

(Dispersion parameter for binomial family taken to be 1)

Null deviance: 85.9502503894332  on 61  degrees of freedom Residual deviance:
                0.0000000030718  on 52  degrees of freedom 
AIC: 20    
Number of Fisher Scoring iterations: 25