set_predictors: Set the predictors of an aba model.

In ncullen93/abaR: Automated Biomarker Analysis

Set the predictors of an aba model.

Description

Predictors are the independent variables which you want to vary as a factor in your statistical models across different groups, outcomes, and stats. Predictors can be supplied as individual variables or as collections of variables, so we refer to a unit of predictors as a "predictor". This function supports both string inputs and actual variables. This function also supports tidy-selection functions like contains and starts_with which allows convenient selection of many variables at once with common names.

Usage

set_predictors(.model, ..., .labels = NULL, .split = FALSE)

Arguments

.model	An aba model. The model for which you want to set predictors
...	strings or variables or tidy-selection functions. Each comma-separated value will be a new predictor set. If you supply actual variables, then the data of the aba model should already be set.
.labels	vector of strings. Optional .labels for printing & plotting. If .labels is set to "self" then the labels will be the predictor values also.
.split	boolean. Whether to split all variables into separate predictors or keep together as one combined predictor. Only relevant when setting predictors with one vector or using things like "starts_with(...)"

Value

An aba model with predictors set.

Examples

data <- adnimerge %>% dplyr::filter(VISCODE == 'bl')

# set with variables - this will result in four "predictor sets".
model <- aba_model() %>%
  set_data(data) %>%
  set_predictors(
    PLASMA_ABETA_bl,
    PLASMA_PTAU181_bl,
    PLASMA_NFL_bl,
    c(PLASMA_ABETA_bl, PLASMA_PTAU181_bl, PLASMA_NFL_bl)
  )

# set with tidy selection functions - but this is only one "predictor set",
# not multiple individual predictor sets.
model <- aba_model() %>%
  set_data(data) %>%
  set_predictors(
    starts_with('PLASMA')
  )

# automatically generate all possible combinations of variables
model <- aba_model() %>%
  set_data(data) %>%
  set_predictors(
    all_combos(c('PLASMA_ABETA_bl', 'PLASMA_PTAU181_bl', 'PLASMA_NFL_bl'))
  )

# supply strings - data does not need to be set first here. But it will
# result in an error if these variables do not éxist in the eventual data.
model <- aba_model() %>%
  set_data(data) %>%
  set_predictors(
    'PLASMA_ABETA_bl',
    'PLASMA_PTAU181_bl',
    'PLASMA_NFL_bl',
    c('PLASMA_ABETA_bl', 'PLASMA_PTAU181_bl', 'PLASMA_NFL_bl')
  )

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