Examples Gallery

Short, focused examples demonstrating TanaT’s core features. Each example is self-contained and can be adapted to your own data.

Data Containers

Build and explore TanaT’s core data structures: event, interval and state sequences, as well as trajectories that combine multiple sequence types for a single individual.

The Three Types of Sequences
Build a SequencePool and explore its contents (event, interval, and state sequences).

Trajectories
Combine several sequences into Trajectory and TrajectoryPool.

Visualisation

Visualize sequences using SequenceVisualizer.

Timeline
Sequences aligned on a shared time axis, flat or category stacking.

Barplot
Count occurrences, relative frequencies, or total durations per label.

Spanplot
Duration distributions as box, violin, or strip plots.

Distribution
State-occupancy areas over time for state sequence pools.

Temporal Alignment

Align sequences to a common reference date (T0 / index date) before computing distances or visualising cross-individual comparisons.

Sequence Zeroing
Define a reference point (T0) for sequences.

Trajectory Zeroing
Define a reference point (T0) for trajectories across multiple sub-pools.

Criteria

Filter sequences and entity rows using composable criteria.

EntityCriterion
Select sequences or prune entity rows with a Polars expression on temporal data.

StaticCriterion
Filter sequences by per-ID static features (age, group, …).

TimeCriterion
Select entities or sequences inside a temporal window; overlap or containment mode.

PatternCriterion
Match ordered patterns of values with free-gap (ANY) and wildcard (*) sentinels.

LengthCriterion
Select sequences by number of entity rows.

RankCriterion
Keep entities at specific positional ranks, absolute or relative to T0.

Metrics

Entity Metrics

Distance metrics for individual entities.

Hamming Distance
Calculate distance between categorical states.

Sequence Metrics

Distance metrics for entire temporal sequences.

LinearPairwise
Position-wise alignment with aggregation.

Edit Distance
Needleman-Wunsch alignment with insertions/deletions.

LCP
Longest Common Prefix distance between sequences.

LCS
Longest Common Subsequence distance.

DTW
Dynamic Time Warping with flexible alignment.

SoftDTW
Differentiable DTW variant.

Chi²
Chi-squared distance between distributions.

Trajectory Metrics

Distance metrics for multi-sequence trajectories.

Aggregation
Compare trajectories with flexible per-type metrics.

Clustering

Partition sequences or trajectories into groups using distance matrices.

Hierarchical Clustering
Nested partitions with dendrograms and four linkage methods.

PAM Clustering
Medoid-based clustering robust to outliers.

CLARA Clustering
Scalable medoid-based clustering for large datasets.