Run dose-response modeling using PROAST.
Usage
f.proast(
odt = list(),
ans.all = 0,
er = FALSE,
resize = FALSE,
scale.ans = FALSE,
const.var = FALSE,
show.warnings = FALSE,
interactive_mode = TRUE,
datatype = NULL,
model_choice = NULL,
setting_choice = NULL,
nested_model_choice = NULL,
indep_var_choice = NULL,
Vyans_input = NULL,
covariates = NULL,
custom_CES = 0.05,
model_selection = NULL,
lower_dd = NULL,
upper_dd = NULL,
selected_model = NULL,
adjust_CES_to_group_SD = NULL,
model_averaging = NULL,
num_bootstraps = NULL,
display_plots = TRUE,
add_nonzero_val_to_dat = FALSE,
nonzero_val = NULL,
detection_limit = NULL
)Arguments
- odt
List. Internal state object/list passed between PROAST functions. Usually, users do not need to set this.
- ans.all
Output from a previous fit, or internal results object. Used to resume or adjust analyses. Usually 0 (default) to start a new session.
- er
Logical. If TRUE, attempt to resume analysis from previously stored state. Used internally/recoverably. Defaults to FALSE.
- resize
Logical. If TRUE, resize the graphics window during execution; passed to graphics helper functions. Defaults to FALSE.
- scale.ans
Logical. If TRUE, applies scaling to the answers/results (advanced use only). Defaults to FALSE.
- const.var
Logical. If TRUE, constrains variance during model fitting (advanced option for troubleshooting). Defaults to FALSE.
- show.warnings
Logical. If TRUE, print extra warning messages during model fitting (for debugging or detailed output). Defaults to FALSE.
- interactive_mode
A TRUE/FALSE value specifying whether you want to run interactively (i.e., TRUE, the default) or using command-line mode (i.e., FALSE, non-interactive). If FALSE, you must provide all other parameters.
- datatype
Non-interactive mode parameter. What type of response data do you want to consider? Options are 'continuous, individual data'.
- model_choice
Non-interactive mode parameter. Do you want to fit a single model or fit various nested families of models? Options are 'single model', 'select model 3 or 5 from various families of models', 'select model 3 from various nested families of models', 'select model 5 from various nested families of models', 'select model 15 in terms of RPF'. Recommended: 'select model 3 or 5 from various families of models'.
- setting_choice
Non-interactive mode parameter. Do you want to fit a set of models, or choose a single model? Options are 'single model', 'set of models'. Recommended: 'set of models'.
- nested_model_choice
Non-interactive mode parameter. Which subset of nested models to fit, if changing model settings non-interactively. Options match those provided in interactive menus. See details in documentation.
- indep_var_choice
Non-interactive mode parameter. The column name for the independent variable to use.
- Vyans_input
Non-interactive mode parameter. The column name(s) for the response variable(s) to use. If multiple, provide as a vector.
- covariates
Non-interactive mode parameter. The column name for the covariate to use. If none, enter 0.
- custom_CES
Non-interactive mode parameter. The critical effect size (BMR) to use, when adjust_CES_to_group_SD = 1 (FALSE).
- model_selection
Non-interactive mode parameter. The model selection to use. Options are "Exponential model only", "Exponential and Hill model", "previous option with inverse exponential model added" (run Expon, Hill, and Inv-Expon), "previous option with lognormal DR model added" (run Expon, Hill, Inv-Expon, and LN). Recommended: "previous option with lognormal DR model added".
- lower_dd
Non-interactive mode parameter. The lower constraint on d parameter. If NULL, existing defaults are used.
- upper_dd
Non-interactive mode parameter. The upper constraint on d parameter. If NULL, existing defaults are used.
- selected_model
Non-interactive mode parameter. Which model do you want to continue with? Options are "exponential", "Hill", "inverse exponential", "lognormal DR". Thefunction will output results for all models regardless of this choice. Really just to bypass the menu option. Recommended: "exponential".
- adjust_CES_to_group_SD
Non-interactive mode parameter. Set the BMR to the group standard deviation. Options are 1 (FALSE) or 2 (TRUE).
- model_averaging
Non-interactive mode parameter. Whether to perform model averaging to calculate 90% confidence intervals. TRUE/FALSE.
- num_bootstraps
Non-interactive mode parameter. The number of bootstraps to perform for model averaging. Recommended: 200.
- display_plots
Non-interactive mode parameter. Whether to display plots. TRUE/FALSE.
- add_nonzero_val_to_dat
Non-interactive mode parameter. When the response data contains 0s, whether to add a non-zero value to each observation. TRUE/FALSE. If TRUE, set the nonzero_val parameter with your desired (positive) number. If FALSE, a detection limit will used. Provide the detection limit in the detection_limit parameter. If no detection_limit is given, the function will use the minimum non-zero value in the data. Values below the detection limit will be plotted as half the detection limit.
- nonzero_val
Non-interactive mode parameter. The non-zero value to add to each observation when add_nonzero_val_to_dat = TRUE. Must be a positive number.
- detection_limit
Non-interactive mode parameter. The detection limit to use when add_nonzero_val_to_dat = FALSE. If NULL, the minimum non-zero value in the data will be used. This paramater accepts a numeric value, which will be applied to all response values, or a column name in the data, which will be used to apply different detection limits to different observations.