EventSequenceData()

Use EventSequenceData() constructors to build event-sequence objects, similar to SequenceData() for state sequences.

Constructor Usage

EventSequenceData.from_tse(
    data=None,
    id=None,
    timestamp=None,
    event=None,
    end_event=None,
    event_labels_order=None
)

EventSequenceData.from_state_sequences(
    seqdata,
    event_representation="transition",
    use_labels=True,
    weighted=True,
    end_event=None,
    event_labels_order=None
)

TraMineR Parameter Mapping

• data / id / timestamp / event -> TraMineR data / id / timestamp / event
• seqdata -> TraMineR seqdata
• event_representation -> TraMineR tevent-like conversion rule
• use_labels -> TraMineR use.labels
• weighted -> TraMineR weighted
• event_labels_order -> TraMineR alphabet

Entry Parameters

Parameter	Required	Type	Description
data	✓*	DataFrame / SequenceData	Main input. Use either this, or provide id + timestamp + event directly.
id	✓*	array-like	Person or case identifier.
timestamp	✓*	array-like	Event time for each row.
event	✓*	array-like	Event label for each row.
end_event	✗	str	Optional event label used as end marker.
event_representation	✗	str / np.ndarray	Conversion rule from state to event sequence. Use a string such as "transition", "state", or "period", or pass an array-like transition-definition matrix.
use_labels	✗	bool	Use state labels when converting from state sequences.
weighted	✗	bool	Keep sequence weights when available.
event_labels_order	✗	list[str]	Optional custom order of event labels.

* Provide either data, or (id, timestamp, event).

What It Does

• Validates required columns and missing values.
• Sorts events by id and time.
• Builds a clean event dictionary and internal event codes.
• Returns an EventSequenceData object used by later functions.

Examples

1. From timestamped event table (recommended)

from sequenzo.event_sequences import EventSequenceData

eseq = EventSequenceData.from_tse(data=tse_df)

2. From vectors

eseq = EventSequenceData.from_tse(
    id=df["id"],
    timestamp=df["timestamp"],
    event=df["event"]
)

3. From state sequences

eseq = EventSequenceData.from_state_sequences(
    seqdata=seqdata,
    event_representation="transition"
)

R Counterpart

• Closest R function: seqecreate
• Mapping note: Sequenzo follows the same core idea: build event-sequence objects from TSE-style input (id, timestamp, event) or from state sequences, via EventSequenceData constructors.

Notes

• Input order does not need to be pre-sorted; the function sorts it for you.
• When event_representation is an array-like transition definition, it plays a role similar to TraMineR's tevent transition matrix; when it is a string such as "state", it uses a built-in conversion rule.
• Event names with (, ) or , may make subsequence string matching harder.

Authors

Code: Yuqi Liang

Documentation: Yuqi Liang

References

Ritschard, G., Bürgin, R., & Studer, M. (2013). Exploratory Mining of Life Event Histories. In J. J. McArdle & G. Ritschard (Eds.), Contemporary Issues in Exploratory Data Mining in the Behavioral Sciences (pp. 221-253). Routledge.