README.md

Sequence Tool Kit

seqtk provides various tools for Sequence.

By using seqtk, you could declare data manipulation logic while deferring the actual data indexing. This could be useful for data loading pipleines of machine learning libraries such as PyTorch's torch.utils.data.

Installation

pip install seqtk

Documentation

Mapping

Returns a mapping of a function over the sequence.

Case1: Single sequence

Signature: seqtk.map(func: Callable[[S], T], sequence: Sequence[S])

Example:

import seqtk
seq = [1, 2, 3]
view = seqtk.map(lambda v: v + 1, seq)
view[0]  # 2

Case2: Multiple sequences

Example (n=2):

import seqtk
from typing import Tuple

def f(u: int, v: str) -> Tuple[int, str]:
  return u, v

seq0 = [1, 2, 3]
seq1 = ["a", "b", "c"]
view = seqtk.map(f, seq0, seq1)
view[0]  # (1, "a")

Gather

Returns a view on the sequence reordered by indices.

Signature: seqtk.gather(sequence: Sequence[T], indices: Sequence[int])

Example:

import seqtk
seq = [1, 2, 3]
view = seqtk.gather(seq, [2, 0])
view[0]  # 3

Concatenate

Returns a view on the concatenated sequences.

Signature: seqtk.concatenate(sequences: Sequence[Sequence[T]])

Example:

import seqtk
seq = [1, 2, 3]
view = seqtk.concatenate([seq, seq])
len(view)  # 6

Development

Test

pytest tests/

Lint

pysen run lint