Generate plots for predicted/observed site level wastewater concentration

[sbidari]

As per the meeting today, order sites by population and display six most populous sites for each state in the PDF that gets reviewed

[damonbayer]

@sbidari You will want to start by modifying this section to properly parse the new variables

pyrenew-hew/hewr/R/process_state_forecast.R

Lines 361 to 368 in 888c04c

	posterior_predictive_variables <-
	pyrenew_posterior_predictive |>
	colnames() |>
	stringr::str_remove("\\[.+\\]$") |>
	unique() |>
	purrr::keep(~ stringr::str_starts(., "observed_")) |>
	stringr::str_c("[group_time_index]") |>
	purrr::map(rlang::parse_expr)

In particular, you may need to identify the multidimensional variables in a different way, since the [group_time_index] will not be valid for identifying [group_time_index, site_index] variables.