Cashflower

Resolved issue causing unnecessary `AttributeError` exceptions to be thrown.

This release adds an option to run only part of the model points from the command line.

Example:

python
python run.py 2 3


This runs the **2nd chunk out of 3**, splitting the model points into equal parts.

Add caching do the `get` method of the `ModelPointSet` class to improve runtime.

Updates:
* The log file name has been changed from `<timestamp>_log.txt` to `<timestamp>.log`.
* Logs will now be saved even if there is an error during execution.

This version introduces significant improvements to make defining model point sets easier and more flexible:

- **No need to define model point sets in `input.py`** - if no model point sets are defined, a single calculation will be performed.
- **Flexible naming** - model point set no longer needs to be named `main`. Instead, for multiple sets, exactly one must have the `main` parameter set to `True` (e.g., `ModelPointSet(main=True)`).
- **Simplified `ID_COLUMN`** - the `ID_COLUMN` setting has been removed. Use the `id_column` parameter directly in the `ModelPointSet` class (e.g., `ModelPointSet(id_column=...`)).
- **Updated template** - the initial template now includes helpful comments to make getting started easier.

These changes reduce complexity.

This release introduces checks to ensure the settings are correct. It verifies that all settings are properly defined and that the output variables correspond to existing variables in the model.