Create a Directed Acyclic Graph (DAG)
create_dag.RdThis function creates a DAG using the DiagrammeR package. It allows for the inclusion of various biases and paths.
Usage
create_dag(
exposure_name,
outcome_name,
label_inquiry = "Causal Inquiry",
confounder_path = NULL,
surrogate_confounder = NULL,
collider_path = NULL,
notoriety_bias = NULL,
drug_competition_bias = NULL,
event_competition_bias = NULL,
background_dilution = NULL,
add_measurements = FALSE,
add_reporting = FALSE,
ascertainment_drug = NA,
ascertainment_event = NA,
additional_rows = "",
rankdir = "LR",
scenario = "inquiry"
)Arguments
- exposure_name
Character. The name of the exposure variable.
- outcome_name
Character. The name of the outcome variable.
- label_inquiry
Character. The label for the inquiry. Default is "Causal Inquiry".
- confounder_path
List. A list containing nodes and signs (and potentially direction) for the confounder path. By deafault it does not show it.
- surrogate_confounder
List. A list containing the surrogate confounder and its root. By deafault it does not show it.
- collider_path
List. A list containing nodes and signs for the collider path. By deafault it does not show it.
- notoriety_bias
Character. The notoriety bias variable. By deafault it does not show it.
- drug_competition_bias
Character. The drug competition bias variable. By deafault it does not show it.
- event_competition_bias
Character. The event competition bias variable. By deafault it does not show it.
- background_dilution
List. A list containing the drug and event for background dilution. By deafault it does not show it.
- add_measurements
Logical. Whether to add measurements. Default is FALSE.
- add_reporting
Logical. Whether to add reporting Default is FALSE.
- ascertainment_drug
Character. Whether to draw an ascertainment bias introduced by the drug. Default is NA.
- ascertainment_event
Character. Whether to draw an ascertainment bias introduced by the event. Default is NA.
- additional_rows
Character. Parameter to add more lines for customised DAGs.
- rankdir
Character. The direction of the graph layout. Default is "LR", or left to right. It can also be set "TB" or top to bottom.
- scenario
Character. The causal scenario type. Default is "inquiry", leaving the question mark on the relation between drug and event. It can also be causal and non-causal.