Function to perform ancestral state reconstruction using corHMM. Output is parsed into a dataset with episodes of trait gain, loss, and continuation across the phylogeny
Usage
asr(
df,
tr,
tip_name_variable,
trait,
model = "ER",
node_states = "joint",
upper_bound = 1e+50,
lower_bound = 1e-09,
confidence_threshold = NULL
)Arguments
- df
Dataframe with tip name (e.g., tip_name_variable) and phenotype/trait (e.g., trait) variables
- tr
Phylogeny object of class "phylo"
- tip_name_variable
Name of variable containing tip names in the dataframe (df). Tip name variable must correspond to tip names in the tree.
- trait
Name of phenotype/trait variable in the dataframe (df).
- model
Type of rate transition matrix. Options: equal rates model ("ER") or all rates different ("ARD"). The option, "MF", selects the best performing model using sample-size corrected Akaike information criterium (AICc). Default: ER
- node_states
Perform "joint" or "marginal" reconstruction. Default: joint
- upper_bound
Upper bound for likelihood search. Default: 1e50
- lower_bound
Lower bound for likelihood search. Default: 1e-9
- confidence_threshold
The confidence threshold to categorize ancestral state inferences from a marginal reconstruction.We suggest using a value of 0.5 (i.e., winner takes all), but permit modification to elevated values. Required when node_states == "marginal".
Value
- corHMM_output
corHMM output
- corHMM_model_statistics
Dataframe containing the inferred rates, log-likelihood, AIC, and chosen transition model. See documentation of characterize_asr_model for more information.
- parent_child_df
Dataframe of parent-child relationships with additional descriptive information
- node_states
Characetr string whether "joint" or "marginal" reconstruction was performed