WOLFRAM

Wolfram 语言与系统参考资料中心
"String" (比较方法)

"String" (比较方法)

Compare two strings using any distance metric.

Details

  • By default, the "String" comparison method considers two strings to be equivalent if they are identical.
  • The distance metric and the corresponding tolerance can be set using the DistanceFunction and Tolerance option. EditDistance is the default distance function.
  • The DistanceFunction is only used for nonzero Tolerance. Other compatible distance metrics include HammingDistance and DamerauLevenshteinDistance.
  • Additionally, options for the distance metrics, like IgnoreCase, are also supported in the second argument of the AssessmentFunction.
  • When multiple values in the answer key are within the tolerance range of an answer, the closest value is selected.

范例

打开所有单元 关闭所有单元

基本范例  (3)

Create an AssessmentFunction for a string question:

Wolfram Language code: stringaf = AssessmentFunction["Dog", "String"]

Use it to assess an answer:

Wolfram Language code: stringaf["Dog"]
Wolfram Language code: stringaf["Dig"]

Define assessment for a DNA sequence with tolerance:

Wolfram Language code: dnatest = AssessmentFunction["AGGTCCCAAAA", Tolerance -> 2]

Any sequence within an EditDistance of two is considered correct:

Wolfram Language code: dnatest["AGGTTCCAAAT"]

Set multiple answer keys with custom scores and infer the string method:

Wolfram Language code: stringaf = AssessmentFunction[{"Dog" -> 2, "Cat" -> -2}]
Wolfram Language code: result = stringaf["Dog"]

Get more details from the result:

Wolfram Language code: result[{"Score", "AnswerCorrect"}]

Scope  (2)

Perform category assessment involving strings:

Wolfram Language code: asmf = AssessmentFunction[{"Iguana" -> <|"Category" -> "Reptile", "Score" -> 1|>, "Moose" -> <|"Category" -> "Mammal", "Score" -> 1|>}]
Wolfram Language code: asmf["Moose" -> "Mammal"]

Assess each element of a list separately:

Wolfram Language code: assessment = AssessmentFunction[{"tall" -> 1, "big" -> 2, "giant" -> 3, "small" -> -1}, <|"ListAssessment" -> "SeparatelyScoreElements"|>][{"big", "small"}]

Get the full result details:

Wolfram Language code: assessment[All]

Options  (2)

Set a custom tolerance:

Wolfram Language code: stringAF = AssessmentFunction[{"Dog" -> 2, "Cat" -> -2}, Tolerance -> 1]
Wolfram Language code: stringAF["Doug"]
Wolfram Language code: stringAF["Drug"]

Compute equivalent comparisons directly:

Wolfram Language code: EditDistance["Doug", "Dog"]
Wolfram Language code: EditDistance["Drug", "Dog"]

Using different distance metrics might produce different results:

Wolfram Language code: AssessmentFunction["Doug", Tolerance -> 2, DistanceFunction -> EditDistance]["Dugo"]
Wolfram Language code: AssessmentFunction["Doug", Tolerance -> 2, DistanceFunction -> HammingDistance]["Dugo"]

This is consistent with using the distance metrics directly:

Wolfram Language code: EditDistance["Doug", "Dugo"]
Wolfram Language code: HammingDistance["Doug", "Dugo"]

Applications  (1)

Define a utility function to make BioSequenceComplement support DNA strings:

Wolfram Language code: flipdna[str_] := BioSequenceComplement[BioSequence["DNA", str]]["SequenceString"]

Create a QuestionObject for determining DNA complements allowing one mistake in each:

Wolfram Language code: QuestionObject[ QuestionInterface["TextCompletion", <| "Prompt" -> "Give the complementary genetic sequences:", "Template" -> "ACTGG `` GACTT ``" |> ], AssessmentFunction[{flipdna /@ {"ACTGG", "GACTT"}}, <|"ComparisonMethod" -> "String", "ListAssessment" -> "AllElementsOrdered"|>, Tolerance -> 1] ]

Properties & Relations  (1)

When an answer matches multiple values in the answer, the closest value is chosen:

Wolfram Language code: AssessmentFunction[{"AAAAAAAB" -> 10, "AAAAABBBB" -> 2}, "String", Tolerance -> 10]["AAAABBB"][All]

Possible Issues  (2)

Some metrics only work when comparing strings with equal length:

Wolfram Language code: AssessmentFunction["Dog", DistanceFunction -> HammingDistance]["Doug"]

The "String" comparison method is based on distances, not patterns. Matching of string patterns is not supported:

Wolfram Language code: AssessmentFunction["A*", "String"]["Apple"]

Instead, specify StringMatchQ as a custom comparator:

Wolfram Language code: AssessmentFunction["A*", StringMatchQ]["Apple"]