main.py

"""
This file contains master functions for the different cryopeg brine scenarios.
Run the method that corresponds to the scenario you'd like to simulate.
All relevant parameters will be calculated for that scenario including:
    - Maintenance energy high and low bounds
    - Model output given the calculated maintenance energy
    - Brine expansion factor
The main function of this file executes all scenarios and the sensitivity analysis.
Call run_all_analysis() with custom parameters to create your own scenario.
"""

import csv
import os
import pickle
from scenario import *
from plots import *


def log_results(analyses, cached_SA):
    """
    Displays or saves the plots for each analysis done. Prints or save endpoints of model and maintenance energy
    calculations. Saves plots and values to disk if savelog is true.
    """
    csv_header = ["Analysis title", "Surrounding pOC", "Brine pOC", "Predicted brine pOC", "Surrounding dOC",
                  "Brine dOC", "Predicted brine dOC", "Present cell density", "Predicted end cell density",
                  "EEA Lower", "EEA Upper", "Real EEA", "EEA predicted timespan", "dOC/cell",
                  "Maintenance energy lower bounds", "Maintenance energy upper bound", "Simulation growth rate",
                  "Minimum growth rate", "Minimum doubling time", "Growth yield", "Brine expansion factor"]
    csv_rows = []

    for analysis in analyses:
        # Endpoints and ME values
        values = [analysis.title,
                  # POC
                  np.format_float_scientific(analysis.scenario.start_poc, precision=2),
                  np.format_float_scientific(analysis.scenario.end_poc, precision=2),
                  np.format_float_scientific(analysis.model_result.pOC[-1], precision=2),

                  # DOC
                  np.format_float_scientific(analysis.scenario.start_doc, precision=2),
                  np.format_float_scientific(analysis.scenario.end_doc, precision=2),
                  np.format_float_scientific(analysis.model_result.dOC[-1], precision=2),

                  # Cell density
                  np.format_float_scientific(analysis.scenario.observed_end_cell_density, precision=2),
                  np.format_float_scientific(analysis.model_result.cells[-1], precision=2),

                  # EEA
                  np.format_float_scientific(analysis.eea_estimation.eea_lower, precision=2),
                  np.format_float_scientific(analysis.eea_estimation.eea_upper, precision=2),
                  np.format_float_scientific(analysis.scenario._eea_rate, precision=2),
                  np.format_float_scientific(analysis.eea_estimation.predicted_timespan/365.25, precision=2),

                  # Cell carbon content
                  np.format_float_scientific(analysis.scenario.dissolved_organic_carbon_per_cell, precision=2),

                  # Maintenance energy
                  np.format_float_scientific(analysis.maintenance_energy_result.lower_bound_me, precision=4),
                  np.format_float_scientific(analysis.maintenance_energy_result.upper_bound_me, precision=4),

                  # Growth rate
                  np.format_float_scientific(analysis.scenario._growth_rate, precision=4),
                  np.format_float_scientific(analysis.maintenance_energy_result.minimum_growth_rate, precision=4),
                  np.format_float_scientific(analysis.maintenance_energy_result.minimum_doubling_time / 365.25, precision=2),

                  # Growth yield
                  np.format_float_scientific(analysis.growth_yield, precision=4),

                  # Brine expansion
                  np.format_float_scientific(analysis.expansion_result.ratio_dimensions, precision=2)]

        csv_rows.append(values)

        # Save plots and write values to CSV
        # Make a plots folder is it doesn't exist
        if not os.path.exists("Results/"):
            os.mkdir('Results/')

        if analysis.sensitivity_analysis_result:
            sa_fig = plot_sensitivity(analysis)
            sa_fig.savefig("Results/" + analysis.title.replace("$", "") + "_sa.tif", format="tif", dpi=300)

            # Save SA values if not from cache
            if not cached_SA:
                with open("Results/SA_result_object", "wb") as sa_results_file:
                    pickle.dump(analysis.sensitivity_analysis_result, sa_results_file)

    # Write the values to CSV if required
    with open("Results/values.csv", "w+") as f:
        write = csv.writer(f)
        write.writerow(csv_header)
        write.writerows(csv_rows)

    # Plot for all scenarios all analyses
    all_analyses_fig = plot_all_scenarios_all_analyses(analyses)
    single_datatype_plot_cells = plot_one_result_type_all_analyses(analyses, "Cells")

    all_analyses_fig.savefig("Results/all_model_outputs.tif", format="tif", dpi=300)
    single_datatype_plot_cells.savefig("Results/all_model_outputs_just_cells.tif", format="tif", dpi=300)

    # Example usage
    interactive_plot = plot_one_result_type_all_analyses_interactive_plotly(all_analyses, "Cells")
    interactive_plot.show()

    return


if __name__ == "__main__":  # Generates all figures and data points.
    cached_results = False
    if not cached_results:
        # All sensitivity analyses should be the same.
        scenarios = [cb1_scenario(), cb4_scenario(), cbiw_scenario()]
        all_analyses = []

        # On every scenario, try every analysis configuration.
        for use_me_lower_bound in [True, False]:
            for scenario in scenarios:
                a = Analysis(scenario, use_minimum_growth_rate=False, use_me_lower_bound=use_me_lower_bound, use_eea_average=False)
                a.run_analysis(do_sensitivity_analysis=False, cached_SA=False)
                all_analyses.append(a)

        # Save plots and values of all results.
        with open("Results/all_analyses", "wb") as aa_results_file:
            pickle.dump(all_analyses, aa_results_file)

        log_results(all_analyses, False)

    else:  # Load results and create plots
        with open("Results/all_analyses", "rb") as aa_results_file:
            all_analyses = pickle.load(aa_results_file)
            log_results(all_analyses, False)