cat_formula_prop_table.py

"""
This script generates multiple statistal analysis tables for fitting cations using all 16 formulas and all 9 properties.
Change variables before `if __name__ == "__main__"` to experiment with different combinations.
Output is a LaTex formatted table, which you can directly copy to LaTex editor.
"""

import modules as mod
from tabulate import tabulate
import pandas as pd
from contextlib import suppress
import time
import os

ion_type = "cations"
file_name = f"csv/{ion_type}.csv"
formula_list = ["Q1R1", "Q1R2", "Q2R1", "Q2R2", "Q3R1","Q3R2", "Q4R1", "Q4R2", "Q5R1", "Q5R2", "Q6R1","Q6R2", "Q7R1", "Q7R2", "Q8R1", "Q8R2"]
property_list = ["dG", "dH", "viscosity", "diffusion", "gfet", "dlcst"]
charge_list = [1, 2]
save_dir_path = "tables"

if __name__ == "__main__":
    start = time.perf_counter()
    if not os.path.exists(save_dir_path):
        os.makedirs(save_dir_path)
    for every, charge in enumerate(charge_list):
        for per, property in enumerate(property_list):
            stats_list = []
            aicc_list = []
            for each, formula in enumerate(formula_list):
                formula_func = getattr(mod.formulas, formula)
                df = mod.read_data(file_name, property, formula, by_charge=False)
                df = df[df.Charge == charge]
                with suppress(TypeError, ZeroDivisionError):
                    prop, fit_data, parameters, SE = mod.fit_parameters(formula, formula_func, property, df)
                    mean_abs_err, rmse, loocv, aicc = mod.get_errors(prop, fit_data, parameters)
                    aicc_list.append(aicc)
                    stats_df = mod.print_stats(formula, df, parameters, SE, mean_abs_err, rmse, loocv, aicc)
                    stats_list.append(stats_df)
                
            stats_result = pd.concat(stats_list, axis=1)
            stats_result = mod.replace_aicc(aicc_list, stats_result)
            stats_result = stats_result.fillna("-")
            with open(f"{save_dir_path}/{ion_type[:3]}{charge}_{property}.txt", "w", encoding="utf-8") as f:
                f.write(f"+{df.iloc[0, 0]} Charge\n")
                f.write(tabulate(stats_result, headers="keys", stralign="center", tablefmt='fancy_grid'))
            print(f"Finished with +{charge}: {property}..")     # It should stop with "ValueError: 'ydata' must not be empty" at +2: dlcst because no values for those ions. 

    end = time.perf_counter()
    print(f'Finished in {round(end-start,2)} seconds') # 13s on gpu desktop.