plot_local_fraud: Plot samples from eforensics
In DiogoFerrari/eforensics:

This function creates a plot with posterior distribution of fraud in a given election unit (ballot box, pooling place, etc)

plot_local_fraud(data, samples = NULL, model = NULL, row = NULL,
  election.unit = NULL, title = NULL, subtitle = TRUE,
  plot.mean = TRUE, legend.position = "bottom", sim_data = NULL,
  by.types = TRUE)

`data`	either the data set used to estimate fraud or the output of the function ef_get_local_fraud
`samples`	eigher `NULL` (default) (which requires `data` to be the output of the function ef_get_local_fraud) or the outuput of the function eforensics
`model`	either `NULL` (default) (which requires `data` to be the output of the function ef_get_local_fraud) or a string with the name of the model used to estimate fraud (see eforensics)
`row`	an integer with the row number of data to plot
`election.unit`	a string with the name of the column that contains a label that identifies the election unit. Default `NULL`
`title`	a string with the title of the plot. Default `NULL`
`subtitle`	a string with the subtitle of the plot. Default `NULL`
`plot.mean`	boolean, if `TRUE` the posterior average of the distribution is also displayed as well as the 95% HPD
`legend.position`	a string with the position of the legend when posterior mean is displayed. Accepts `top`, `bottom`, `left`, `right`
`sim_data`	the output of the function ef_simulateData. It is used only if the fraud was estimated using simulated data. If provided, the plot also display the true fraud value. Default `NULL`
`by.types`	boolean, if `TRUE` incremental, extreme, and total fraud are displayed. Otherwise, only distribution of total fraud is shown