customDataGen()
takes in an R data frame and processes it to generate data suitable for the
drawr()
function.
Usage
customDataGen(
df,
xvar = NULL,
yvar = NULL,
regression_type = "linear",
success_level = NULL,
degree = NULL,
span = 0.75,
log_y = F,
log_base = NULL,
conf_int = F
)
Arguments
- df
An R data frame containing the input data.
- xvar
The name of the x variable as a string. If null is provided will use first column of dataframe (default: NULL)
- yvar
The name of the y variable as a string. If null is provided will use second column of dataframe. (default: NULL)
- regression_type
Type of regression data to generate. Options include "linear", "polynomial", "logistic" (only for binary logistic regression), or "loess". (Default: "linear")
- success_level
Which level of binary categorical variable should be considered success if using "logistic" regression. If NULL uses alphabetical order. (Default: NULL)
- degree
The degree for a polynomial or loess regression. If chosen in regression_type. For loess 'degree' must be 0, 1 or 2. (Default: 2 for polynomial, 1 for loess)
- span
The span for a loess regression. (Default: 0.75)
- log_y
Specify whether to apply a logarithmic transformation to y for the fitted line for when using a non-linear scale in the
drawr()
function. Currently only for linear regression. If TRUE, the fitted line is transformed as log(y) ~ x; if FALSE, the fitted line is not transformed. (default: FALSE)- log_base
Log base for log transformation, only applies if log_y is true. If NULL will apply a natural log transformation. Log_base should match log_base choice in
drawr()
function. (default: NULL)- conf_int
If a 95% confidence interval should be generated for
drawr()
function. Currently only for linear regression. (default: FALSE)