Add column to toyMC results with minuit convergence flag

[kdund]

This PR adds the minuit convergence flag to the results.
Closes #90

Example to check:

from alea.utils import load_yaml
from alea.runner import Runner
import numpy as np

runners = dict()
for fittable_sigma in [False, True]:
    runner = Runner(
        statistical_model = "alea.examples.gaussian_model.GaussianModel",
        poi = "mu",
        hypotheses = ["free","zero","true"],
        n_mc = 100,
        nominal_values = {"sigma":1.0},
        generate_values = {"mu":1.0},
        parameter_definition= {
        "mu": {
            "fit_guess": 0.0,
            "fittable": True,
            "nominal_value": 0.0,
            "parameter_interval_bounds": [
                -10,
                10,
            ],
        },
        "sigma": {
            "fittable": fittable_sigma,
            "nominal_value": 1.0,
            "fit_limits": [0,np.inf],
        },
    },
    output_filename = "test_sigmafree_{:d}.hdf".format(int(fittable_sigma))
    )
    runner.run()
    
    

for sigma_free in [False, True]:
    run_results = ii.toyfiles_to_numpy("test_sigmafree_{:d}.hdf".format(int(sigma_free)))
    print([np.mean(run_results[k]["valid_fit"]) for k in ["free","zero","true"]]) #the second, only 0.25 of fits converge or so

[dachengx]

Looks cool. Thanks @kdund .