Dissimilarity Helper Functions

Most users should start with get_distance_matrix(). It provides the main public interface for sequence dissimilarities.

This page documents helper functions that are useful when you need to inspect cost matrices, reproduce a TraMineR-style calculation, or build custom diagnostics around a distance method.

When to Use Helpers

Need	Helper
Build a substitution-cost matrix	get_substitution_cost_matrix()
Inspect common-prefix length for two sequences	get_LCP_length_for_2_seq()
Inspect OMstran backend construction	build_omstran_substitution_matrix()
Compute chi-square Euclidean distances	chi2_euclid_distances()
Convert a list of sequence rows to a distance-ready form	lower-level utilities in sequenzo.dissimilarity_measures

Substitution-Cost Matrices

from sequenzo import get_substitution_cost_matrix

sm = get_substitution_cost_matrix(seqdata, method="CONSTANT", cval=2)

Substitution-cost matrices are passed to optimal-matching methods through get_distance_matrix() or to lower-level alignment diagnostics such as pairwise_sequence_alignment().

from sequenzo import get_distance_matrix, pairwise_sequence_alignment

diss = get_distance_matrix(seqdata, method="OM", sm=sm, indel=1)
alignment = pairwise_sequence_alignment(seqdata, indices=[0, 1], indel=1, sm=sm)

Prefix Helpers

Prefix-based distances such as LCP and suffix-based variants such as RLCP are usually computed through get_distance_matrix(). If you need the common-prefix length for a pair of sequences, use:

from sequenzo import get_LCP_length_for_2_seq

length = get_LCP_length_for_2_seq(seq_a, seq_b)

For full prefix/suffix tree analysis, use the Prefix and Suffix Trees page.

Specialized Distance Backends

Some distance routines exist as lower-level backend exports because they are useful for validation, benchmarking, or method development. They are not the recommended user-facing route for ordinary analyses.

from sequenzo.dissimilarity_measures.measures_implemented_with_python import (
    build_omstran_substitution_matrix,
    chi2_euclid_distances,
)

In normal documentation examples, prefer:

diss = get_distance_matrix(seqdata, method="OMstran")

or the corresponding method name in get_distance_matrix() when it is available.

build_omstran_substitution_matrix() is an internal-facing backend helper. It expects intermediate OMstran objects such as transition-state data and precomputed substitution/indel costs, so most users should call get_distance_matrix(..., method="OMstran") instead.

Professional Workflow Guidance

• Use get_distance_matrix() for production analyses.
• Use helper functions when you need to expose an intermediate object, such as a substitution-cost matrix.
• Keep the cost matrix, indel cost, normalization, and missing-value handling together in your project notes; these choices are part of the statistical specification, not only coding details.
• When validating against TraMineR, compare both the distance matrix and the cost matrix used to produce it.

See Also

• get_distance_matrix() is the main entry point these helpers support.
• Matrices in Dissimilarity Measures explains the matrices involved.

Authors

Code: Yuqi Liang