CharBudgetCompressor#

class scikitplot.corpus.CharBudgetCompressor(max_chars, whole_words=True)[source]#

Keep each hit’s text, best first, until a character budget is spent.

Parameters:
max_charsint

Characters of source text the context may hold in total.

whole_wordsbool, default=True

Cut a partly kept hit at its last whitespace character (str.isspace) within the budget, so no word is split; a hit whose first word does not fit keeps nothing. Text without whitespace (Chinese, Japanese) is then kept only whole: use False for it.

Parameters:

  • max_chars (int)

  • whole_words (bool)

Notes

User. Deterministic and order-preserving: hits are kept in rank order and each keeps one span from the start of its text. Hits after the budget is spent keep nothing.

Examples

>>> class Doc:
...     def __init__(self, text):
...         self.text = text
>>> class Hit:
...     def __init__(self, text):
...         self.doc = Doc(text)
>>> CharBudgetCompressor(12)("q", [Hit("alpha beta"), Hit("gamma delta")])
[((0, 10),), ()]
__call__(query, hits)[source]#

Return one tuple of spans per hit.

Parameters:
Return type:

list[tuple[tuple[int, int], …]]

max_chars: int[source]#
whole_words: bool = True#