Random-forest-mc

1) Add the different Python versions in the automated tests.

1) Add `NaN`, `None`, `Null` cheking before use and generate the Tree.
2) Remove `self.dataset = dataset.dropna()` (l. 509, `model.py`).

1) fix: `survived_score` getting wrong value. When the Tree is not dropped, the value got is `survived_score` and not th_val. Is not crictical, because generally the trees require many trees, so the score is got correctly.
2) Add quality checks in `utils.LoadDicts`.
3) Add input parameter `ignore_errors` in `utils.LoadDicts`.
4) Add iterable protocol in `utils.LoadDicts`.
5) Refactor some type hints.
6) Add missing values prediction feature.

1) Add the parameter `max_depth` to limit how deep the branching will be.
2) Add the `min_samples_split`, once get this minimun or less, the leaf is generated instead a new branching.
3) Fix the `__let__` that was with `>` instead `<`.
4) Remove the threading mode for fit in paralell processing.

1) Replace sum with fsum in many parts., from math standard lib. See: https://docs.python.org/3/library/math.html.