FeatureNearest[{elem1,elem2,…},x]
给出在算出的特征空间中,x 为其最近元素的 elemi 列表.
FeatureNearest[{elem1v1,elem2v2,…},x]
给出与 x 为其最近元素的 elemi 对应的 vi.
FeatureNearest[{elem1,elem2,…}{v1,v2,…},x]
给出同样结果.
FeatureNearest[{elem1,elem2,…}prop,x]
给出 x 为其最近元素的 elemi 的属性 prop.
FeatureNearest[data,{x1,x2,…}]
实际上给出 {FeatureNearest[data,x1],FeatureNearest[data,x2],…}.
FeatureNearest[data,x,n]
给出离 x 最近的 n 个 elemi.
FeatureNearest[data]
生成 NearestFunction[…],可以重复用于不同的 x.
FeatureNearest
FeatureNearest[{elem1,elem2,…},x]
给出在算出的特征空间中,x 为其最近元素的 elemi 列表.
FeatureNearest[{elem1v1,elem2v2,…},x]
给出与 x 为其最近元素的 elemi 对应的 vi.
FeatureNearest[{elem1,elem2,…}{v1,v2,…},x]
给出同样结果.
FeatureNearest[{elem1,elem2,…}prop,x]
给出 x 为其最近元素的 elemi 的属性 prop.
FeatureNearest[data,{x1,x2,…}]
实际上给出 {FeatureNearest[data,x1],FeatureNearest[data,x2],…}.
FeatureNearest[data,x,n]
给出离 x 最近的 n 个 elemi.
FeatureNearest[data]
生成 NearestFunction[…],可以重复用于不同的 x.
更多信息和选项
- FeatureNearest 在自动计算的特征空间中求包括文本、图像、声音和数值数据在内的多种数据类型最近的元素.
- FeatureNearest 可被用于许多不同类型的数据,包括数值、文字、声音、图像,以及上述类型的组合.
- 每个 elemi 可以是单个数据元素、数据元素列表、数据元素关联或 Dataset 对象. 在 FeatureNearest[data,…] 中,data 可以是 Dataset 对象.
- 在 FeatureNearest[{elem1,elem2,…}prop,…] 中,prop 的可能形式包括:
-
"Element" 找到的最近的 elemi "Index" 找到的最近的 elemi 的索引 i "Distance" 到最近的 elemi 的距离 {prop1,prop2,…} 多种形式的列表 All 给出元素、索引和距离的关联 - 当返回几个元素时,先给出最近的一个.
- FeatureNearest 通常等价于在由 FeatureExtract 生成的空间中使用 Nearest.
- 可以给出下列选项:
-
DistanceFunction Automatic 要使用的距离度量 FeatureExtractor Identity 怎样提取特征 FeatureTypes Automatic 假设的输入数据的特征类型 Method Automatic 要使用的方法 RandomSeeding 1234 应该在内部对伪随机数生成器进行什么样的初始化 - RandomSeeding 的可能设置包括:
-
Automatic 每次函数调用时自动重新播种 Inherited 使用外部播种的随机数字 seed 用明确给定的整数或字符串作为种子 - Method 的可能设置包括 "Octree"、"KDtree" 和 "Scan".
范例
打开所有单元 关闭所有单元基本范例 (1)
用 FeatureNearest 产生最近函数:
nf = FeatureNearest[{{1.4, "A"}, {1.5, "A"}, {2.3, "B"}, {5.4, "B"}}]nf[{1.7, "A"}]FeatureNearest[{{1.4, "A"}, {1.5, "A"}, {2.3, "B"}, {5.4, "B"}}, {1.7, "A"}]范围 (5)
nf = FeatureNearest[{"the cat is grey", "my cat is fast", "this dog is scary" , "the big dog"}]nf[{"what a cat", "this is not a dog"}]在 DateObject 列表上训练最近函数:
nf = FeatureNearest[{DateObject[{2014, 5, 5}, TimeObject[{9, 53, 6.30158}]], DateObject[{2000, 1, 1}, TimeObject[{0, 0, 0.}]], DateObject[{2006, 12}], DateObject[{2007, 8, 23}, CalendarType -> "Gregorian"], DateObject[{2016, 4, 4}, TimeObject[{15, 59, 18.2738}, TimeZone -> -4.], TimeZone -> -4.]}]找出距新的 DateObject 最近的两个元素:
nf[DateObject[{2003, 1, 2}], 2]nf["2nd of January 2003", 2]nf = FeatureNearest[{{"the cat is grey", [image]}, {"my cat is fast", [image]}, {"this dog is scary", [image]}, {"the big dog", [image]}}]nf[{"the cat cat", [image]}]利用输出属性 All 训练最近函数:
nf = FeatureNearest[{[image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image]} -> All]Dataset[nf[[image], 6]]FeatureNearest[{[image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image]}, [image], 3]选项 (4)
DistanceFunction (1)
用距离函数 EditDistance 在文字列表上生成最近函数:
nf = FeatureNearest[{"the cat is grey", "my cat is fast", "this dog is scary" , "the big dog"}, DistanceFunction -> EditDistance]nf[{"what a cat", "this is not a dog"}]在距离函数被指定的情况下,Nearest 和 FeatureNearest 相互等价:
nf2 = Nearest[{"the cat is grey", "my cat is fast", "this dog is scary" , "the big dog"}, DistanceFunction -> EditDistance]nf2[{"what a cat", "this is not a dog"}]FeatureExtractor (1)
用 FeatureExtractor 训练最近函数,以便对数据进行预处理:
nf = FeatureNearest[{DateObject[{2014, 5, 5}, TimeObject[{9, 53, 6.30158}]], DateObject[{2000, 1, 1}, TimeObject[{0, 0, 0.}]], DateObject[{2006, 12}], DateObject[{2007, 8, 23}, CalendarType -> "Gregorian"], DateObject[{2016, 4, 4}, TimeObject[{15, 59, 18.2738}, TimeZone -> -4.], TimeZone -> -4.]}, FeatureExtractor -> AbsoluteTime]nf[DateObject[{2003, 1, 2}], 2]FeatureTypes (1)
用 FeatureTypes 训练最近函数来指定 "gender" 是一个名义变量:
nf = FeatureNearest[{
<|"age" -> 32, "gender" -> 1|>,
<|"age" -> 41, "gender" -> 2|>,
<|"age" -> 17, "gender" -> 2|>,
<|"age" -> 11, "gender" -> 1|>},
FeatureTypes -> <|"gender" -> "Nominal"|>
]nf[<|"age" -> 20, "gender" -> 1|>]应用 (2)
dataset = {[image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image], [image]};用 FeatureNearest 来训练最近函数:
nf = FeatureNearest[dataset]nf[{[image], [image], [image]}]alice = ExampleData[{"Text", "AliceInWonderland"}];sentences = TextSentences[alice];nearestalice = FeatureNearest[sentences]nearestalice["Alice and the Rabbit"]nearestalice["Alice and the Queen"]nearestalice["Off her head"]nearestalice["The Hatter"]文本
Wolfram Research (2017),FeatureNearest,Wolfram 语言函数,https://reference.wolfram.com/language/ref/FeatureNearest.html (更新于 2017 年).
CMS
Wolfram 语言. 2017. "FeatureNearest." Wolfram 语言与系统参考资料中心. Wolfram Research. 最新版本 2017. https://reference.wolfram.com/language/ref/FeatureNearest.html.
APA
Wolfram 语言. (2017). FeatureNearest. Wolfram 语言与系统参考资料中心. 追溯自 https://reference.wolfram.com/language/ref/FeatureNearest.html 年
BibTeX
@misc{reference.wolfram_2026_featurenearest, author="Wolfram Research", title="{FeatureNearest}", year="2017", howpublished="\url{https://reference.wolfram.com/language/ref/FeatureNearest.html}", note=[Accessed: 15-September-2026]}
BibLaTeX
@online{reference.wolfram_2026_featurenearest, organization={Wolfram Research}, title={FeatureNearest}, year={2017}, url={https://reference.wolfram.com/language/ref/FeatureNearest.html}, note=[Accessed: 15-September-2026]}