New function spread_draws for creating tidybayes compliant MCMC output

[SamuelBrand1]

This PR proposes a new function: spread_draws

Rt-without-renewal/EpiAware/src/utilities.jl

Lines 236 to 255 in 048fd0a

	"""
	spread_draws(chn::Chains)
	Converts a `Chains` object into a DataFrame in `tidybayes` format.
	# Arguments
	- `chn::Chains`: The `Chains` object to be converted.
	# Returns
	- `df::DataFrame`: The converted DataFrame.
	"""
	function spread_draws(chn::Chains)
	df = DataFrame(chn)
	df = hcat(DataFrame(draw = 1:size(df, 1)), df)
	@rename!(df, $(".draw") = :draw)
	@rename!(df, $(".chain") = :chain)
	@rename!(df, $(".iteration") = :iteration)
	return df
	end

This converts a Chains object return from running Bayesian inference using the Turing sample method into a DataFrame which conforms to the standard output expected from the tidybayes function of the same name.

Added dependencies:

• DataFramesMeta. This exports DataFrames functionality as well as additional macros for dplyr-like interaction with dataframe objects. In particular, the constructor DataFrame dispatches on Chains objects to create a DataFrame that is nearly tidybayes compliant.
• CSV. This exports utilities for using .csv files. This has been added to the /test folder environment to expose CSV.write.

Closes #55

[seabbs]

Is this intensionally missing things from the test environmment? Like StatsPlots?

[SamuelBrand1]

StatsPlots is in the test env but not a dependency of EpiAware; which I think is the right move.

[seabbs]

This looks good. We might need an issue to resolve the test environment once we sort out documentation as there are things like StatsPlot that don't appear.