CHExNet: A network of collaborators in early modern Jagiellonian University

authority_file_cac_alma.parquet: Parquet table of person authority records (ALMA) with crosswalk to CAC. One row per person (final_id); includes alma_name (authority heading), alma_born/alma_died (years, may be missing/NaN), cac_id (linked CAC identifier, optional), and first_polish_pub (earliest Polish publication year, optional).

ALMA_matched.pkl is a pickled Python object (typically a list[dict]) containing enriched ALMA bibliographic records with normalized titles, date ranges, places, and reconciled contributor/subject entities.

Each element (row) is one ALMA record with these fields (as observed):

• ALMA_id (str): ALMA MMS/bib identifier for the record.

• autors_list (list[str] | None): list of main author name strings (may be missing). 

• title (str | None): main title. 

• subtitle (str | None): subtitle (may be missing). 

• publication_city (list[str] | None): place statements as extracted strings (often “a: … | d: …”). 

• date_start (datetime.date | None): normalized start date for the record’s date span. 

• date_end (datetime.date | None): normalized end date for the record’s date span. 

• record_type (str | None): high-level type label (e.g., “Tekst”, “Obraz”). 

• all_names (numpy array of object | None): all person-name strings associated with the record (authors + contributors + other roles), stored as a NumPy array. 

• all_names_final_id (tuple[int] | None): aligned tuple of IDs corresponding to all_names (same order) and matcheable with authority file's final_id. 

• contributing_persons (list[tuple[str, list[str] | None]] | None): contributor persons as (name, roles) pairs, where roles is a list of role labels (e.g., ['Oprac.'], ['Redakcja']) or None. 

• contributing_organizations (list[tuple[str, str | None]] | None): contributor organizations as (org_name, role) pairs (role may be None). 

• genre (list[str] | None): genre/form terms (can contain duplicates). 

• subjects_people (list[str] | None): subject people headings (optional). 

• subjects_topics (list[str] | None): topical subject headings (optional). 

• subjects_places (list[str] | None): geographic subject headings (optional). 

CAC_matched.pkl is a pickled Python object (typically a list[dict]) containing normalized person-event records from CAC, with standardized place fields, date ranges, and a compact metadata dictionary describing the event.

Each element (row) represents one event instance with these fields (as observed):

• person_id (int): internal person identifier (CAC person/entity ID).

• event_place (str | None): event place label (human-readable).

• event_place_parent (str | None): parent place/institution label (human-readable) - this is the value used for co-presence.

• event_place_id (float | int | None): normalized place identifier for event_place (often float due to missing values/NaNs).

• event_place_parent_id (float | int | None): normalized place identifier for event_place_parent.

• date_start (pandas Timestamp | None): normalized start date (inclusive) for the event time span.

• date_end (pandas Timestamp | None): normalized end date (inclusive) for the event time span.

• event_metadata (dict | None): event-specific attributes; keys vary by event type, commonly:

  • event_type (str): e.g., “uzyskanie stopnia”, “koniec/ustanie funkcji”

  • degree_type (str, optional): e.g., “magister”

  • science_field (str, optional): e.g., “sztuki wyzwolone/filozofia”

  • position (str, optional): e.g., “prowizor”

time_id.pkl is a pickled Python dictionary that assigns a unique integer ID to each discrete time slice (half-year steps).

For more information on how to use the files, see luizdovalle2/CHExNET---Analysis and the article CHExNet: A 400-years Multilayer Network of Early Modern Collaboration at the Jagiellonian University