Error when "# Test our predict_model() function"

[aephidayatuloh]

I try to reproduce the work using code from customer_churn.R. But when I ran this block of code bellow (line 150-152), I get the error

> # Test our predict_model() function
> predictions <- predict_model(x = model_keras, newdata = x_test_tbl, type = 'raw') %>% tibble::as_tibble()
Error in `dimnames<-.data.frame`(`*tmp*`, value = list(n)) : 
  invalid 'dimnames' given for data frame

Can you help me to figure out way to solve this?
Thank you

[neural-oracle]

Hi ... I had same error ... solution to your problem - if still relevant or for future users - should be this one:
thomasp85/lime#139
seems that keras package has new nomen clature ... thus either run the customer churn code with old version of keras and tensor flow ... or update the code by inserting the suggested ... sequential.... as shown under the link above