Using factor inputs in a caret-driven shiny app - shiny

Question

I am trying to develop a simple shiny app which takes in patient data and predicts the probability of a disease condition using caret.
Diagnosis
Age
Creatine
Chronic
70
765
Chronic
80
784
Chronic
72
692
Chronic
88
965
Chronic
68
1065
Chronic
75
1005
Acute
56
445
Acute
67
378
Acute
78
501
Acute
45
678
Acute
37
776
Acute
39
644
The following code works and returns a probability value.
library(shiny)
library(caret)
library(readxl)
hepC_lite <- read_excel("hepC_lite.xlsx")
model_mars <- train(Diagnosis ~ ., data = hepC_lite, method = "earth")
ui <- fluidPage(
numericInput("age", label = "Age", value = 50, min = 1, max = 99),
numericInput("creatine", label = "Creatine", value = 100, min = 1, max = 2000),
actionButton("submitButton", "Submit"),
tableOutput("userDefinedTable"),
textOutput('probability')
)
server <- function(input, output) {
values <- reactiveValues()
observeEvent(input$submitButton, {
values$new_row <- data.frame(Age = input$age, Creatine = input$creatine)
values$predicted_mars <- predict(model_mars, values$new_row, type="prob")[,2]
})
output$userDefinedTable <- renderTable(values$new_row)
output$probability <- renderText(values$predicted_mars)
}
shinyApp(ui = ui, server = server)
To include a factor variable (Gender) in the prediction, I am first one-hot encoding the dataset.
Diagnosis
Age
Creatine
Gender
Chronic
70
765
M
Chronic
80
784
M
Chronic
72
692
F
Chronic
88
965
M
Chronic
68
1065
M
Chronic
75
1005
F
Acute
56
445
F
Acute
67
378
F
Acute
78
501
F
Acute
45
678
M
Acute
37
776
M
Acute
39
644
F
#one-hot encoding
x = hepC_lite[, 2:4]
y = hepC_lite$Diagnosis
dummy_model <- dummyVars(Diagnosis ~ ., data = hepC_lite)
trainData <- predict(dummy_model, newdata = hepC_lite)
hepC_lite <- data.frame(trainData)
hepC_lite$Diagnosis <- y
How should I edit the following line to include the Gender variable?
values$new_row <- data.frame(Age = input$age, Creatine = input$creatine)
Running this line with Gender = input$gender causes an error - Warning: Error in eval: object 'Gender.F' not found