Estimate Dirichlet parameters for samples from multiple sources — fun_estimate_Dirichlet_from

Estimate MLE parameters from a dataframe of Dirichlet samples from different sources. The sources must be known.

fun_estimate_Dirichlet_from_samples(df, col_source = "source", ...)

Arguments

df	dataframe of Dirichlet samples with a source column
col_source	the column name containing the source column
...	Arguments passed on to `fun_estimate_Dirichlet_from_single_source` `name_param` name of output parameter (default: `'theta'`) `use` if `'ML'`, use MLE estimator, else the sample estimators (unbiased for \(\theta\), but not optimal)

dataframe of Dirichlet samples with a source column

col_source

the column name containing the source column

...

name_param: name of output parameter (default: 'theta')
use: if 'ML', use MLE estimator, else the sample estimators (unbiased for \(\theta\), but not optimal)

a tibble containing the Dirichlet parameter estimates for each source

Suppose that rows in source i are \(X ~ Dir(\theta_i)\) iid. This function estimates \(\theta_i\).