Generate chain of events for individuals inc. for mni dataframe

[marghe-molaro]

This PR updates previous PR draft #1468

Now includes changes in the mother_and_newborn_info dataframe.

[marghe-molaro]

Hi @tamuri,

• Logging fixed as suggested
• EventName error fixed
• Utils fncs updated

For now "final package" offered by util fnc includes , for each draw, person ID (combined across runs for that draw), date, event, and attribute changes incurred by individual as part of that event. I think this is how most users will think about processing this info, but we can maybe discuss at Q4.

Note that I am still getting a Run Checks error but this is not linked to ruff:

ERROR: Could not install packages due to an OSError: [Errno 28] No space left on device

[marghe-molaro]

Still to finalise:

• Are the copies of pop dataframe/rows "deep" enough? Do columns containing series and lists need to be copied iteratively?
• Are nans and NaTs handled appropriately? Attributes that are 'nan' before/after event seem to be incorrectly detected as having changed
• Currently saving values as strings to deal with different types, but will save different types to different V columns as previously suggested by Asif (V_bool, V_int, V_float, etc)

[marghe-molaro]

To add following discussion at Q4:

• Add option for users to specify characteristics of individuals for which histories are tracked;
• Ensure events after death are not logged (a few of these may slip through via pop-wide events)

[tamuri]

Are the copies of pop dataframe/rows "deep" enough? Do columns containing series and lists need to be copied iteratively?

The following population dataframe columns have object type i.e. the ones to check are copied/checked properly:

>>> [col for col in df.columns if df[col].dtype == 'object']
['rt_injuries_to_cast', 'rt_injuries_for_minor_surgery', 'rt_injuries_for_major_surgery', 'rt_injuries_to_heal_with_time', 'rt_injuries_for_open_fracture_treatment', 'rt_injuries_left_untreated']

[marghe-molaro]

Hi @tamuri,
This is now ready for review. Notable changes:

• Fixed issue with nan values being incorrectly logged as having changed;
• Unified approach to copying pop df + mni dictionary, and ensured object types are copied by value;
• In postprocessing, ensured events which may have been logged after the individual's death are not kept in the final histories (there won't be many)

[marghe-molaro]

Added features:

• Individuals now have an additional property called track_history. Events are only logged for individuals for whom track_history = True either before or after the event has ran. Developers are free to specify how this variable is updated by events in the simulation.

Note:

• Inpatient_Care is currently not an event associated with a specific individual, therefore (by definition) it is not logged by the individual history tracker

[marghe-molaro]

Now additionally tracking consumable, equipment, and bed days access for each individual