"gdgtBay" offers an accessible opportuntiy to calibrate and predict 
brGDGT distributions to/with climate conditions. The function runs 
on the Bayesia Model-Building Interface, Bambi (Capretto et al., 2022),
which sits above the Bayesian statistical modelling package, PyMC
(Salvatier et al., 2016). In the combination, the user can select and 
modify all bayesian parameters such as prior distributions, sampling
methods (see bambi and PyMC packages for further informatio). 

Each step of the multi-bayesian approach, as outlined by Dearing 
Crampton-Flood et al. (2020), is labelled below. The user has the option
to run a calibration model (predictor_data defulat: none) or calibrate
and then predict results (provide the predictor data of interest; e.g.,
MBT'5me values).

Calibration model - Outputs the results of the the second bayes as well
                    as a figure depicting the true vs. estimated
                    predictand values. To select, use predictor_data
                    default (None).
                  - Useful for exploring the model and testing parameters
                    prior to predicting values.
                  
Prediction model  - Outputs the original predictor_data dataframe with two appended
                    columns, the mean and standard deviation of the predicted 
                    values.


Variable descriptions:
    • gdgt            - Name ofindex or fractional abundance of interest. If 
                        predicting, ensure that predictor_data is lablled 
                        accordingly.
    • clim            - Name of predictand. Must be in data
    • data            - Calibration dataset - must inculde 'gdgt' and 'clim'.
                        Must be in the pandas dataframe format.
    • bayes_summary   - Summary of first bayesian method results.
                        defaults to show.
    • figurepath      - Pathway to save location of figures
    • user_priors     - If true, user can alter priors in function. If false,
                        bambi will select priors.
    • bayes_1_figure  - Figure of posterior predictive ouput from
                        first bayes application. useful for assessing
                        model applicability. Defaults to show and save.
    • true_est_figure - Figure showing true predicted values (x axis) vs
                        estimated values (y axis). Default to show. Only 
                        works for calibraiton-only applications.
    • regression_fig  - Figure showing bayesian regression model (blue) as well
                        as OLS regression model (fuchsia) and calibratio data
                        (black).
    • chains          - Number of Markov chains to run. _# denotes # instance
                        of Bayes. Default as none to have bambi select. See
                        bambi documentation for more details.
    • draws           - Number of draws from sampling method. _# denotes #
                        instance of Bayes. Default as none to have bambi select. 
                        See bambi documentation for more details.
    • cores           - Number of cores for CPU to run. _# denotes #
                        instance of Bayes. Default as none to have bambi select. 
                        See bambi documentation for more details.
    • tune            - Number of samples used to tune predictio model (second 
                        Bayes instance). Default as none to have bambi select. 
                        See bambi documentation for more details.
    • predictor_data  - Optional dataset cotaining predictor (gdgt) data that
                        from which the user wants to estimate some climate variable
                        (clim). Must be a pandas dataframe with the predictor column
                        labelled with the same name as 'gdgt' variables. Feel free to
                        include any other informatio in this dataframe - the prediction
                        model appends the mean and standard deviation of predicted
                        values onto the end of the predictor_data dataframe.
Function depedencies:
    • Bambi
    • PyMC
    • sklearn
    • numpy
    • pandas
    • matplotlib
    """