1

I got this code to perform an internal validation of a binary Logistic Regression model from this -https://rdrr.io/cran/GmAMisc/man/modelvalid.html- and it works perfectly with the dataset suggested data(log_regr_data).

Unfortunately when I use my dataset (10 variables 1 Outcome and 10 dependent variables (9 continuous and 1 categorical)

model <- glm(admit ~ plat,.......... data = data, family = "binomial")
summary(model)

res <- modelvalid(data=data, fit=model, B=100)""

it doesn't work at all and I receive the following error:

Error in roc_(data, response_name, predictor_name, ret = ret, ...) : Column data[, 1] not present in data data Inoltre: Warning message: In roc.data.frame(data[, 1], data$pred.prob.full, data = data) : This method uses a non-standard evaluation (NSE). Did you want to use the roc_ function instead?

This happens even if I use the same name for the dataset. I cannot understand this problem.

Can someone help me?
Thanks in advance

MLavoie
  • 9,671
  • 41
  • 36
  • 56
Fabrizio
  • 11
  • 1
  • Did you check whether your data match the requirements of the function? When it comes to the dataset, in the dataframe containing the dataset the dependent variable must be stored in the first column to the left. – NewAtGis May 16 '21 at 15:44

0 Answers0