aba_evaluate: Evaluate a fitted aba model on new data

In ncullen93/abaR: Automated Biomarker Analysis

Evaluate a fitted aba model on new data

Description

This function lets you evaluate the performance of a fitted aba model on new data.

Usage

aba_evaluate(model, data)

Arguments

model	a fitted aba model. The model to test on new data
data	dataframe. The new data on which the fitted model will be tested.

Value

an object with class 'abaEvaluate' which contains model performance on the data

Examples

# create separate training and testing datasets
data <- adnimerge %>% dplyr::filter(VISCODE == 'bl')
train_id <- sample(nrow(data), 0.8*nrow(data), replace = FALSE)
data_train <- data %>% dplyr::filter(dplyr::row_number() %in% train_id)
data_test <- data %>% dplyr::filter(dplyr::row_number() %in% train_id)

# fit model on train data
model <- aba_model() %>%
  set_data(data_train) %>%
  set_groups(
    everyone(),
    DX_bl %in% c('MCI', 'AD')
  ) %>%
  set_outcomes(CDRSB_bl, MMSE_bl) %>%
  set_predictors(
    PLASMA_PTAU181_bl, PLASMA_NFL_bl,
    c(PLASMA_PTAU181_bl, PLASMA_NFL_bl)
  ) %>%
  set_covariates(AGE, GENDER, EDUCATION) %>%
  set_stats('lm') %>%
  fit()

# evaluate fitted model on test data
res <- model %>% aba_evaluate(data = data_test)

ncullen93/abaR documentation built on Feb. 8, 2024, 12:01 p.m.